2.4. X-rays
The X-ray properties of bright, optically selected quasars have been intensively studied in the last 25 years (Elvis et al. 1978; Zamorani et al. 1981), mostly with broadband but low resolution spectra. The X-ray emission from quasars extends from the Galactic absorption cut-off at ~ 0.1 keV up to ~ 300 keV. Laor et al. (1997) analyzed the ROSAT soft X-ray (0.5 - 2 keV) observations of the sample of local (z < 0.4) PG quasars, and a subsample of these objects has been studied with ASCA in the 2 - 10 keV band (George et al. 2000) and with BeppoSAX in the 1 - 100 keV band (Mineo et al. 2000). Recent studies of samples of bright Seyfert 1 galaxies are reported in George et al. (1998; ASCA observations) and in Perola et al. (2002; BeppoSAX observations). The main properties of the X-ray spectra of type I AGN are briefly summarized below and are shown in Figure 3.
 |
Figure 3. Average total spectrum (thick
black line) and main components (thin grey lines) in the
X-ray spectrum of a type I AGN. The main primary continuum component is
a power law with an high energy cut-off at E ~ 100 - 300 keV, absorbed
at soft energies by warm gas with NH ~
1021 -1023 cm-2. A cold reflection
component is also shown. The most relevant narrow feature is the iron
K |
Primary
emission.
The intrinsic continuum X-ray emission of quasars is to first
order a power law, extending from about 1 keV to over 100 keV.
However, as higher resolution and better signal-to-noise spectra
have become available, emission and absorption features have been
found that mask a direct view of this "power law" over virtually
the whole X-ray band (see Fig. 3).
Hence, slight curvatures may be present but unseen. The typical
spectral index (3) is
between 
= -0.8 and
= -1, both for low
luminosity Seyfert galaxies and high luminosity quasars.
Radio-loud AGN have a somewhat flatter spectrum
(
between -0.5 and -0.7). This is thought to be due to the
additional hard component emitted by inverse-Compton scattering
of the electron in the jet on the radio-synchrotron photons,
but this is not fully established.
There is now increasing evidence, mostly from BeppoSAX, for a roughly exponential cut-off to the power law at energies ~ 80 - 300 keV. This is presumably due to the cut-off in the energy distribution of the electrons responsible for the X-ray emission. It is still debated whether the spectral index is redshift or luminosity dependent (Zamorani et al. 1981; Avni & Tananbaum 1982, 1986; Bechtold et al. 2003; Vignali et al. 2003).
In addition to the main power law continuum component, a soft emission component is often observed in AGN, with characteristic temperature kT ~ 0.2 - 1 keV. The physical origin of this component is not clear: warm emitting gas could be located in the accretion disk, or in the broad line region (it could be the confining medium of the broad emission line clouds), or in a region farther from the center. Alternatively, this "soft excess" could be an extension of the big blue bump to higher energies, e.g., via Compton scattering in a hot accretion disk corona (Czerny & Elvis 1987).
Reflection
components.
The primary emission of AGN can be "reflected", i.e., Thomson
scattered by ionized gas. If the reflector has a column density
NH > 1.5 × 1024 cm-2
(i.e.,
~ 1 / 
T)
and is not fully ionized, the reflected component has a spectrum
like the one shown in Figure 3 (the actual
shape slightly varies, depending on the geometry and chemical
composition of the reflector). The main features of this reflection
component are a continuum due to electron scattering
with a peak at ~ 30 keV, and a cut-off at 4 - 5 keV due to
photoelectric absorption of the lower energy incident radiation.
The reflection efficiency is typically a few percent of the direct
emission in the 2 - 10 keV range because of photoelectric
absorption, rising to ~ 30% at the 30 keV peak for a Compton-thick
reflector covering a significant fraction of the solid angle
(Ghisellini et
al. 1994).
The efficiency drops if the reflecting
medium is Compton thin (in this case part of the incident radiation
escapes without interaction).
A warm, ionized reflector must be present in the central region of many AGN (since we see a "warm absorber" in ~ 50% of Seyfert 1 galaxies). The reflected emission has the same spectral shape as the incident continuum.
Iron line.
The most prominent narrow feature in the 2 - 10 keV X-ray spectra
of AGN is an iron emission line at energy 6.4 keV, corresponding to
the Fe-K n = 2 - 1 transition of "cold" (i.e.,
FeXVII) iron.
The line is usually ascribed to emission due to fluorescence in the
inner part of the accretion disk. The typical EW of
the line is 100 - 200 eV. There is also evidence for a broad "red wing"
extending to lower energies
(Tanaka et al. 1995;
Nandra et al. 1997).
Once thought to be widespread, XMM-Newton spectra now show signs
of this red wing only in a few objects (MCG-6-30-15 being the clearest example,
Vaughan & Fabian
2004;
see Fig. 4).
This red wing has caused great excitement as a likely physical cause
is the gravitational redshift and relativistic Doppler shift of an
Fe-K line originating from an accretion disk at only a few
Schwarzschild radii (RS) from the central black hole.
Such a broad line would be one of the best tools to look for general
relativistic effects in strong gravity. Asymmetric profiles have been
calculated for lines emitted at a few Schwarzschild radii from non-rotating
(Fabian et al. 1989)
and rotating
(Laor 1991)
black holes.
 |
Figure 4. Iron line profile in an XMM-Newton observation of the Seyfert 1 galaxy MCG-6-30-15. Model is a power law fitted in the 3 - 5 keV and 8 - 10 keV energy bands. Figure is from Vaughan & Fabian (2004; their Fig. 8). |
A second, narrow component of the Fe-K line is very clearly present in most AGN. The width of this "narrow" line (which is unresolved in CCD spectra from ASCA, Chandra ACIS, or XMM-Newton EPIC) is a few 1000 km s-1 or smaller when measured with the Chandra HETG spectrograph. This is similar to the width of optical and UV broad emission lines. This narrow component does not vary when the continuum varies, even for delay times of days. Coupled with the line width, this suggests an origin well beyond a few RS, although a small radius is not fully ruled out (Fabian et al. 2002). In the next few years, the ASTRO-E2 satellite, with 6 eV resolution (R = 1000) and an effective area of ~ 150 cm2, is expected to do much better in understanding this issue. If the reflector is highly ionized, the peak energy of this line can be shifted toward high energies (6.7 keV for helium-like iron and 6.96 keV for hydrogen-like iron). It is also possible that two narrow components are present in the spectrum, one emitted by a cold reflector and the other by an ionized reflector. CCD detectors like the ASCA SIS or XMM-Newton EPIC (with energy resolution of ~ 120 - 150 eV at 6 keV) are unable to separate these two lines, while the Chandra HETG spectrograph has limited effective area (~ 40 cm2) at 6 keV.
Warm absorbers.
Warm absorber features are present in the soft X-ray spectra of half
of the bright Seyfert 1 galaxies observed with ASCA
(Reynolds et al. 2000).
Recently, the availability of high resolution
soft X-ray spectra, obtained with the grating instruments onboard
Chandra and XMM-Newton, show that this component is
formed by an outflowing gas. We show in Figure 5 the
highest signal-to-noise high resolution spectrum of an AGN, obtained
with a long observation of the Seyfert 1 galaxy NGC 3783. Recently,
Krongold et al. (2003)
were able to reproduce all of the observed lines
with a two-phase absorber, with the two phases in pressure equilibrium.
 |
Figure 5. Chandra grating spectrum of NGC 3873, superimposed with a simple two-component model fitting most of the absorption features. Figure is from Krongold et al. (2003; their Fig. 3). |
3 X-ray astronomers tend to use the
"photon index"
, where
N(E) 
E- and
= -
( - 1).
Back.