Very-high and very-low luminosity AGN

>From the observations discussed above, we can deduce that the accretion disks in most Seyfert-1 nuclei are only modestly ionized and remain optically-thick (and capable of producing iron line features) down to the radius of marginal stability. Seyfert-1 nuclei provide a useful reference against which we can compare and contrast the accretion disks of other types of AGN. By making such comparisons, we can hope to learn how the properties and physics of accretion disks change as a function of the gross observables of the system. Here, we examine constraints on the dependence of the accretion disk properties on the overall radiative luminosity of the AGN.

We begin by discussing low-luminosity AGN (LLAGN), roughly defined as sources with a total radiative luminosity of $L<10^{42}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$. Being low-luminosity, even the nearest examples are comparatively faint X-ray sources. Furthermore, stellar processes in the galaxy (such as X-ray binaries and vigorous star-formation) can produce a comparable X-ray flux. This galactic emission can be spatially blended with the LLAGN emission and somewhat hamper studies.

One of the best studied LLAGN is the nearby galaxy NGC 4258 (M 106)¹³ -- this is the same object for which masers give an excellent constraint on the black hole mass. The first good constraints on the iron line in this source were obtained with a deep ASCA observation [246] (although previous shorter observations had hinted at such a line [247]). It was found that the iron line was narrow, implying that the bulk of the fluorescent emission originated from $r>100M$. Given the fairly large equivalent width of the line ($\sim 100{\rm\thinspace eV}$) and the lack of any evidence (from any waveband) for cold, optically-thick material away from the plane of the accretion disk, this emission line very likely originates from the accretion disk itself. Thus, to produce such a narrow line, the X-ray emitting corona must be extended, producing appreciable X-ray illumination of the disk across a range of radii exceeding $r>100M$. This is in stark contrast to the higher-luminosity Seyfert galaxies whose broad lines require a very compact X-ray emitting corona ($r<10M$). This result is consistent with a sphere+disk model for this LLAGN with a transition radius of $r_{\rm tr}>100M$. However, due to the limited signal-to-noise, the presence of a weaker relativistic iron line in addition to the narrow component could not be ruled out. Recently it has been reported that a rather short XMM-Newton observation of NGC 4258 failed to detect any iron line, setting an upper limit on the line flux to be lower than that detected by ASCA [248]. Such variability suggests that the iron line does indeed originate from the accretion disk (rather than more distant material) and, furthermore, implies that the disk/corona system undergoes significant changes in structure and/or ionization state on year time scales.

X-ray studies of the general population of LLAGN have been hampered by their comparative faintness. The best such study to date, based upon ASCA data, detects significant iron line emission in several LLAGN [249,250]. In no object, however, is the line emission found to be significantly broadened. Unfortunately, the signal-to-noise of these datasets is insufficient to draw robust astrophysical conclusions -- relativistic iron lines such as those seen in higher-luminosity Seyfert nuclei could readily be hiding in the noise of these data. However, there are excellent prospects for characterizing any broad line component with current and future XMM-Newton and Chandra observations. Such observations will allow us to compare the properties of accretion disks (especially ionization state and truncation radii) in LLAGN with those in normal Seyfert nuclei.

We now turn to a brief discussion of high-luminosity AGN (HLAGN). Again, due to the maturity of the data, much of our understanding is still largely based on ASCA observations. In an important study, Nandra and co-workers studied the average ASCA spectrum of a large sample of AGN as a function of source luminosity ranging from normal Seyfert nuclei to powerful quasars [251]. They found that broad iron line becomes appreciably weaker once one considers sources with an X-ray luminosity greater than $L_X\sim 10^{44}-10^{45}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$. The most reasonable interpretation for this trend is that the HLAGN possess more highly ionized accretion disks -- this is expected if the luminosity of an AGN is determined primarily by the Eddington fraction rather than the black hole mass. Support for this picture is growing with subsequent XMM-Newton observations. XMM-Newton studies of the high-luminosity Seyfert galaxies Mrk 205 and Mrk 509 (which possess an X-ray band luminosity of $\sim 10^{45}\hbox{${\rm\thinspace erg}{\rm\thinspace s}^{-1}\,$}$) find significant broadened K$\alpha$ iron lines, with profiles well described with accretion disks models, and energies corresponding to helium- or hydrogen-like iron [252,253]. These observations likely represent direct detections of ionized accretion disks.

Direct evidence for ionized disks have also been found in the X-ray spectra of the so-called narrow-line Seyfert-1 galaxies (NLS1s, [254,255]). NLS1s are a peculiar sub-class of Seyfert-1 galaxies which are defined by their rather narrow optical emission lines, together with very soft and highly variable X-ray emissions [256]. While they do not possess a high luminosity in absolute terms, it is thought that NLS1s may have rather small black hole masses and luminosities rather close to their Eddington limit. The detection of highly ionized disks in these objects nicely fits in with this scenario.