• Kinkhabwala, A.
• Kaastra, Jelle
• Branduardi-Raymont, G.
• Page, M. J.
• Paerels, F.
• Kahn, S. M.
• Liedahl, D. A.
• Sako, M.
• Behar, E.
• Brinkman, A. C.
• Den Herder, J. S.

Abstract

The interpretation of the soft X-ray spectra of the Seyfert 1 galaxies, MCG-6-30-15 and Mrk 766, has remained controversial since high resolution data were first obtained with the grating instruments on Chandra and XMM-Newton, roughly one year ago. In an initial paper, Branduardi-Raymont et al. (2001), we argued that the RGS spectra of these two sources are inconsistent with simple warm absorber models, as has been invoked for Seyfert 1s in the past, but instead suggest the additional presence of relativistically broadened disk line features associated with the Lyα transitions of carbon, nitrogen, and oxygen. This conclusion was subsequently questioned by Lee et al. (2001), who contended that the Chandra HETG spectrum of MCG-6-30-15 is indeed well-described by the conventional warm absorber model, if one allows for the presence of dust in the warm absorbing medium. Here we reexamine the original RGS spectra in light of the Lee et al. (2001) criticisms. We first show that the explicit model presented by Lee et al. (2001) for MCG-6-30-15 is incompatible with the RGS data on this source, even if we allow both the continuum parameters and all of the absorbing column densities to be free parameters. That model over-predicts the ion{O}{VII} absorption line equivalent widths, and yields significant systematic residuals to the fits, especially at longer wavelengths, beyond the band covered by the HETG. We next show that the column densities of the oxygen ions (ion{O}{IV} through ion{O}{VIII}) are very well-constrained by the absorption line structure in the RGS data, and that, contrary to the assertion by Lee et al. (2001), the derived values are much too low to provide any significant contribution (either from line or continuum absorption) to the observed discrete jump in the spectra near 17.5 Å. Further, we show that the RGS spectra are also incompatible with the dust model presented by Lee et al. (2001). Specifically, the derived upper limit on the neutral oxygen column density is nearly a factor of 35 lower than predicted by their model if the dust is in the form of simple iron oxides. If dust is indeed present in the warm absorber, it would have to be essentially in the form of pure iron to be compatible with the soft X-ray spectrum. In contrast, a model that includes the presence of relativistically broadened CNO Lyα lines, provides an excellent description of the data, correctly accounting for all of the discrete warm absorber lines plus the overall continuum shape for both sources. We suggest that these emission lines are produced via recombination in a photoionized layer on the surface of an irradiated accretion disk. The derived equivalent widths are roughly compatible with what we expect for this layer if one takes proper account of the modification to the disk structure due to the irradiation, and of continuum and line opacity in the ionized surface layer. <P />...

Topics

• density
• impedance spectroscopy
• surface
• Carbon
• Oxygen
• Nitrogen
• iron