The IwasawaTaniguchi effect for radioquiet AGN

1
The Iwasawa-Taniguchi effect for radio-quiet AGN
X-ray surveys, Rodos Island, July 3, 2007
Giorgio Matt
(Dipartimento di Fisica, Università Roma Tre,
Italy) Stefano Bianchi (RM3), Matteo
Guainazzi and Nuria Fonseca Bonilla
(ESAC/ESA)
2
The Iwasawa-Taniguchi (a.k.a. X-ray Baldwin)
effect
Iwasawa Taniguchi (1993) found an
anti-correlation between the 6.4 keV iron Ka
emission line and the 2-10 keV luminosity, based
on Ginga observations of 37 AGN. This
'Iwasawa-Taniguchi effect' was then found in a
larger sample of objects observed by XMM-Newton,
giving a relation between Lx and EW of the
(narrow core) of the Fe Ka EW Lx (-0.170.08)
(Page et al. 2004).
Page et al. (2004)
Iwasawa Taniguchi (1993)
3
The significance of this effect was questioned
by Jiménez-Bailón et al. (2005) in their analysis
of XMM-Newton data of PG quasars, pointing out
the importance of contamination from radio-loud
objects Jiang et al. (2006), using also Chandra
data, confirmed the IT effect and suggested that
the anti-correlation can be attributed to
variations of the continuum, if the iron line
does not respond istantaneously (i.e. if
emitted from distant matter)
4
(Zhou Wang 2005)
The anti-correlation of the iron line EW with the
Eddington ratio may be stronger than with
luminosity (Zhou Wang 2005)
5
The XMM-Newton Catalogue of Radio-Quiet AGN

• All Radio-Quiet AGN in targeted XMM-Newton p-n
  observations in the public archive with
• gt 200 counts in both 0.5-2 and 2-10 keV (rest
  frame) bands
  
• lt 1 pileup
• NH lt 2x1022 cm-2
• log(R)lt1 (quasar) log(R)lt2.4 and
  log(RX)lt-2.755 (Seyfert)
• 157 sources

All spectra reanalysed - If more than one obs.
per source is available, the longest is chosen -
6 and 20 cm fluxes for the whole sample - BH mass
for 52 of the sources - L-dependent bolometric
correction adopted (Marconi et al. 2004)
6
Narrow core only !!
The IT effect is highly significant 157 data
points (81 measures, 76 upper limits) ?-0.33 -
P4x10-5
7
The anti-correlation of the EW with the Eddington
ratio is highly significant 82 data points (50
measures, 32 upper limits) ?-0.38
P6x10-4 (Still significant if a constant
bolometric correction is used)
log(EWFe)(1.610.05) (-0.190.05) log(Lbol/Edd)
8
No significant dependence of the iron EW on the
BH mass is found 82 data points (50 measures, 32
upper limits) ?-0.15 P0.3
log(EWFe)(1.730.04) (-0.070.04) log(LBH,8)
9
Luminosity-dependent covering factor of the torus?

• Most of the iron line flux comes from the
  torus, not the BLR
• (e.g. Bianchi et al. 2004, Nandra 2006), and must
  originate in
• optically thick matter (NHgt1023 cm-2)
• The most obvious explanation is in terms of
• a luminosity-dependent covering factor of the
  torus.
• This is in agreement with models which predict an
  increase
• of the opening angle of the torus with luminosity
  
• (e.g. Konigl Kartje 1994)
• Interestingly, this implies a decrease of the
  fraction of
• obscured AGN with luminosity as observed (Ueda
  et al. 2003
• La Franca et al. 2005), albeit in Compton-thin
  sources

10
This is in agreement with the recent discovery of
a non-linear relation between thermal emission
from dust and optical luminosity in AGN (Maiolino
et al. 2007), which implies a decrease of the
covering factor of dust with luminosity, with a
slope of -0.18 (-0.17 for the IT effect) .
Maiolino et al. ( 2007)

• Caveat
• The agreement is very good and deserves further
  investigation, even if the two methods may sample
• different materials
• - (Almost) Compton-thick gas for the iron EW
• (NHgt 1023 cm-2)
• - Dust associated both to Compton-thin and thick
  gas for the IR luminosity

11
Variability effects?
Continuum variability naturally produces an
anticorrelation between X-ray luminosity and iron
line EW, and definitely contributes to the IT
effect. However, the simulated anticorrelation
has a slope of -0.050.05 (Jiang et al. 2006),
much weaker than the observed IT effect.
The amplitude of variability in radio-quiet AGN
is generally smaller than one order of magnitude,
while the IT effect is observed on six orders of
magnitude in luminosity, making this explanation
unlikely to explain the whole effect.
NGC 3783 (Jiang et al. 2006)
12
Is the ionization state of the torus increasing
with luminosity?
The correlation between the ratio of highly
ionized iron to neutral iron with luminosity is
rather weak
Bianchi et al. ( 2007)
13
Is the IT effect due to a decrease of the optical
depth of the torus with luminosity? If so, a
steeper effect should be found for the
reflection component.
Simbol-X simulations for a constant EW/R ratio
(as for C-thick matter)
Matt et al. (2003)
14
Summary and outlook
Based on a large sample of XMM-Newton
observations, we confirmed the Iwasawa-Taniguchi
(a.k.a. X-ray Baldwin) effect for radio-quiet
AGN, i.e. the anticorrelation between the EW of
the narrow core of the iron line and the X-ray
luminosity. The effect seems to be mainly due to
an anticorrelation with the Eddington
ratio. The most likely explanation is a decrease
with luminosity (Eddington ratio) of the covering
factor of the torus (similar effect found in IR)
Future To increase the XMM-Newton sample
-- To search for the IT effect in the reflection
component (Simbol-X)