Testing the Unified Model with X-ray Spectra

Keeping in mind a possible modification of the unification model, it would be worthwhile to learn more about the material causing the obscuration in the Schmitt et al. sample. For this purpose, X-ray spectroscopy provides us with a unique diagnostic tool. Seyfert galaxies are defined to be type 2 if there is obscuring material that is optically thick to visual light. However, the same column of material will not be optically thick for photons of sufficiently high energy. Direct nuclear emission can and often is observed in the X-ray waveband. Therefore, we can study the intervening material by studying the X-ray spectral features of Seyfert galaxies. We propose to examine the X-ray spectra of the galaxies in the Schmitt et al. sample in order to place constraints on the location and orientation of the absorbing material and to further probe their hypothesis.

Our sample consists of the 23 of the 46 galaxies in the Schmitt et al. sample for which X-ray spectra are available from the ASCA (Advanced Satellite for Cosmology and Astrophysics) archive. ASCA has a bandpass of 0.5-10 keV. With an energy resolution of 2-4% around 6.4 keV, it has a higher energy resolution than contemporaneous X-ray observatories.

Our plan is to fit models of the following basic form:

absorber1 ( spectrum1 + absorber2 × spectrum2 ) where spectrum1 is the emission from the warm scattering region, spectrum2 is the direct emission of the AGN engine plus fluorescent Fe K emission, absorber2 is the absorbing column which lies along the line of sight to the AGN engine but does not cover the scattering region (ie- the putative torus), and absorber1 is the absorption common to both spectral components. The latter includes absorption in our own galaxy (which is a known quantity) plus any absorption that takes place internally within the source galaxy.

Overall, we will keep the spectral model components as simple as possible. Spectrum1 will be fit with a power law. Spectrum2 includes a power law continuum component and an emission feature for the iron fluorescence line. We will also consider adding a model component for reflection off of a cold, dense cloud. This would modify the flux entering the absorbing column of the torus by assuming there are two contributors to the incident light; one representing direct illumination, and the other with a spectrum modified by Compton scattering and atomic features representing the light reflected off the opposite side of the inner torus edge. In this case, spectrum2 will be modeled as follows:

spectrum2 = powerlaw + iron emission + cold, dense cloud reflection We will only use the reflection component if it offers a significant improvement to our fit. We will ultimately measure the hydrogen column density (N_H) of the two absorbers, the spectral index of the X-ray continuum, and the energy and equivalent width of the Fe K- line.

Photons in the range of 2-10 keV are not completely blocked by photoelectric absorption until N_H nears  [Kallman and Mushotzky, 1985]. We can therefore use the X-ray spectrum to measure column densities for absorber2 up to this level. By fitting for two values of N_H, we will be able to discriminate between absorption which is due to localized obscuration such as the torus and that which is global due to the host galaxy. We expect that type 1 Seyferts will have a small N_H, and the other Seyfert types will have a range of values for the tori column densities. (Note that each of these values is in fact a lower limit for the thickness of each torus, because we can not directly constrain its orientation. The best we could do is use the standard assumption that the torus axis is parallel to the observed radio axis, but we would not know the angle of projection between this axis and the plane of the sky.)

Our goals are (1) to compare the spectral indices of the different types, and (2) to compare the intrinsic X-ray power of both Seyfert types. This will tell us whether there are any systematic differences between the type 1 and type 2 AGN engines. We will also (3) determine the energy of the Fe K- line, which will indicate the ionization state of the gas producing it. Finally, we will (4) measure the equivalent width of the fluorescence line, which we will combine with the measured N_H values and the theoretical models of Ghisellini [Ghisellini et al., 1994] and Krolik [Krolik et al., 1994] in order to constrain the density, the Fe abundance, and the inclination of the torus relative to the orientation of the galaxy and the accretion disk.