A Decade of Rossi X-ray Timing Explorer Seyfert Observations:An RXTE Seyfert Spectral Database

Abstract

With over forty years of X-ray observations, we should have a grasp on the X-ray nature of active galactic nuclei (AGN). The unification model of Antonucci and Miller (1985) offered a context for understanding observations by defining a "typical" AGN geometry, with observed spectral differences explained by line-of-sight effects. However, the emerging picture is that the central AGN is more complex than unification alone can describe. We explore the unified model with a systematic X-ray spectral study of bright Seyfert galaxies observed by the Rossi X-Ray Timing Explorer (RXTE) over its first 10 years. We develop a spectral-fit database of 821 time-resolved spectra from 39 Seyfert galaxies fitted to a model describing the effects of an X-ray power-law spectrum reprocessed and absorbed by material in the central AGN region.

We observe a relationship between radio and X-ray properties for Seyfert 1s, with the spectral parameters differing between radio-loud and radio-quiet Seyfert 1s. We also find a complex relationship between the Fe K equivalent width (EW) and the power-law photon index (Gamma) for the Seyfert 1s, with a correlation for the radio-loud sources and an anti-correlation for the radio-quiet sources. These results can be explained if X-rays from the relativistic jet in radio-loud sources contribute significantly to the observed spectrum. We observe scatter in the EW-Gamma relationship for the Seyfert 2s, suggesting complex environments that unification alone cannot explain. We see a strong correlation between Gamma and the reflection fraction (R) in the Seyfert 1 and 2 samples, but modeling degeneracies are present, so this relationship cannot be trusted as instructive of the AGN physics. For the Seyfert 1 sample, we find an anticorrelation between EW and the 2 to 10 keV luminosity (L_X), also known as the X-ray Baldwin effect. This may suggest that higher luminosity sources contain less material or may be due to a time-lag effect. We do not observe the previously reported relationship between Gamma and the ratio of L_X to the Eddington luminosity.