Introduction

The X-ray study of Seyfert nuclei and other types of active galactic nuclei (AGN) has been energized for the past few years by the observation of relativistically broad iron K lines in their X-ray spectra (Tanaka et al. 1995; Nandra et al. 1997; Reynolds 1997). In particular, the Seyfert galaxy MCG-6-30-15 has become an important testing ground for models of broad iron line formation. A long observation of MCG-6-30-15 by the Advanced Satellite for Cosmology and Astrophysics (ASCA) revealed a high signal-to-noise broad iron line with a velocity width of and a profile which is skewed to low energies (Tanaka et al. 1995). The excitement stirred by these studies is due to the widely held belief that the iron lines originate from the surface layers of an accretion disk which is in orbit about a supermassive black hole, and that the line width and profile provide a direct probe of the velocity field and strong gravitational field within a few Schwarzschild radii of the black hole. Models of line emission from the inner regions of a black hole accretion disk (e.g., Fabian et al. 1989; Laor 1991) fit the observed line profiles well.

The suggestion that we are observing the immediate environment of an accreting supermassive black hole is a bold one and certainly warrants a critical examination. In this spirit, Fabian et al. (1995; hereafter F95) examined a number of alternative hypotheses for the origin of these broad iron lines including models in which the line is produced in an outflow or jet, and models in which the line is intrinsically narrow (or even absent) and a complex underlying continuum mimics the broad line. Both of these classes of models were found to be unphysical or did not reproduce the observed spectrum.

Another alternative model, first proposed by Czerny, Zbyszewska & Raine (1991) but also considered by F95, is one in which the iron line is intrinsically narrow (i.e., emitted in slowly moving material which is very far from a compact object) and then broadened to the observed profile by Compton downscattering in matter that surrounds the source of line photons. F95 rejected this model on the basis that the Comptonizing cloud must have a radius of in order to maintain the required high ionization state and that, with such a small radius, gravitational effects from a central black hole would be important anyway for determining the line profile. The principal aim of F95 was to demonstrate the need to include strong gravity in any model of the iron line, so they terminated their chain of reasoning at that point. The question remains, however, as to whether Compton downscattering has a significant affect on the line profile or whether we can interpret iron line observations in terms of naked accretion disk models.

Misra & Kembhavi (1998) and Misra & Sutaria (1999; hereafter MS99) have recently developed the Comptonization model further. In their current model, they suggest that a cloud with optical depth and temperature surrounds the central engine. The upper limit to the temperature of the Compton cloud comes from the fact that the iron K line photons need to be primarily downscattered, rather than upscattered, in order to reproduce the observed line profile. The central engine produces the continuum emission which keeps the cloud ionized, and a narrow iron line which is Compton broadened to the observed width. They show that the resultant line profiles can be brought into good agreement with the ASCA observations.

A direct prediction of the Comptonization model (F95, MS99) is that the multiple Compton scatterings should produce a break in the spectrum of the power-law continuum radiation at approximately (i.e., - ). Recently, it has been reported that BeppoSAX (Guainazzi et al. 1999) observations constrain the location of the continuum break to be at energies greater than , thereby arguing against the Comptonization model (Misra 1999). However, a robust determination of the continuum break is not completely straightforward since it depends upon the parameters assumed for the Compton reflection component (e.g., see Lee et al. 1999). Thus, while the lack of a spectral break at 30-40keV remains the most compelling argument against the Comptonization model, it is interesting to consider constraints on the Comptonization model that are independent of a continuum spectral break.

In this paper, we apply a number of observational constraints to the MS99 model. We focus on the case of the iron line in MCG-6-30-15, but also address the line in NGC 3516, the other high signal-to-noise case of a relativistically broad line. We show that the continuum source in MCG-6-30-15 required by the constrained model violates thermodynamic limits (i.e., the ``black body'' limit). We also show that only a very small region of parameter space is open to the Comptonization model in the case of NGC 3516. Hence, we conclude that the Compton downscattering model is not a viable model for the broad iron lines in one, and possibly both, of these sources.