High Energy spectra of Seyferts and Unification schemes

The Unified Model of AGN predicts the sole difference between Seyfert 1 and Seyfert 2 nuclei is the viewing angle with respect to an obscuring structure around the nucleus. High energy photons above 20 keV are not affected by this absorption if the column is Compton thin, so their 30--100 keV spectra should be the same. However, the observed spectra at high energies appear to show a systematic difference, with Seyfert 1's having $Γ\sim $2.1 whereas Seyfert 2's are harder with $Γ\sim$ 1.9. We estimate the mass and accretion rate of Seyferts detected in these high energy samples and show that they span a wide range in $L/L_{Edd}$. Both black hole binary systems and AGN show a correlation between spectral softness and Eddington fraction, so these samples are probably heterogeneous, spanning a range of intrinsic spectral indices which are hidden in individual objects by poor signal-to-noise. However, the mean Eddington fraction for the Seyfert 1's is higher than for the Seyfert 2's, so the samples are consistent with this being the origin of the softer spectra seen in Seyfert 1's. We stress that high energy spectra alone are not necessarily a clean test of Unification schemes, but that the intrinsic nuclear properties should also change with $L/L_{Edd}$