2011MNRAS.413.1206B

We present an X-ray spectral analysis of 126 galaxies of the 12 µm galaxy sample. By studying this sample at X-ray wavelengths, we aim to determine the intrinsic power, continuum shape and obscuration level in these sources. We improve upon previous works by the use of superior data in the form of higher signal-to-noise ratio spectra, finer spectral resolution and a broader bandpass from XMM-Newton. We pay particular attention to Compton thick active galactic nucleus (AGN) with the help of new spectral fitting models that we have produced, which are based on Monte Carlo simulations of X-ray radiative transfer, using both a spherical and torus geometry, and taking into account Compton scattering and iron fluorescence. We use this data to show that with a torus geometry, unobscured sightlines can achieve a maximum equivalent width of the Fe Kα line of ∼150 eV, originally shown by Ghisellini et al. In order for this to be exceeded, the line of sight must be obscured with N_H> 10²³ cm^–2, as we show for one case, NGC 3690. We also calculate flux suppression factors from the simulated data, the main conclusion from which is that for N_H ≥ 10²⁵ cm^–2, the X-ray flux is suppressed by a factor of at least 10 in all X-ray bands and at all redshifts, revealing the biases present against these extremely heavily obscured systems inherent in all X-ray surveys. Furthermore, we confirm previous results from Murphy & Yaqoob that show that the reflection fraction determined from slab geometries is underestimated with respect to toroidal geometries. For the 12 µm selected galaxies, we investigate the distribution of X-ray power-law indices, finding that the mean (#915;= 1.90^+0.05_–0.07 and σ_Γ= 0.31^+0.05_–0.05) is consistent with previous works, and that the distribution of Γ for obscured and unobscured sources is consistent with the source populations being the same, in general support of unification schemes. We determine a Compton thick fraction for the X-ray AGN in our sample to be 18±5 per cent which is higher than the hard X-ray (>10 keV) selected samples. Finally we find that the obscured fraction for our sample is a strong function of X-ray luminosity, peaking at a luminosity of ∼10^42–43 erg.s^–1.