Ni%20ABUNDANCE%20IN%20THE%20CORE%20OF%20THE%20PERSEUS%20CLUSTER:%20AN%20ANSWER%20TO%20THE%20SIGNIFICANCE%20OF%20RESONANT%20SCATTERING%20AND%20SNIa%20ENRICHMENT

1
Ni ABUNDANCE IN THE CORE OF THE PERSEUS CLUSTER
AN ANSWER TO THE SIGNIFICANCE OF RESONANT
SCATTERING AND SNIa ENRICHMENT

• Fabio Gastaldello (CNR-IASF, University of
  Milan-Bicocca)
• Silvano Molendi (CNR-IASF)

2
OUTLINE

• RESONANT SCATTERING
• THE QUESTION RS OR INCREASED Ni ABUNDANCE IN THE
  CORE OF THE PERSEUS CLUSTER ?
• THE XMM OBSERVATION OF PERSEUS AND HIS ANSWER TO
  THE QUESTION
• CONCLUSIONS AND SUMMARY

3
RESONANCE SCATTERING
Absorption of a line photon followed by immediate
reemission in an other direction. Although for
densities and temperatures typical of clusters
the gas is optically thin to Thomson scattering
for the continuum, it can be optically thin in
the resonance X-ray lines (in particular in the
denser core) and in particular in the Fe He ?
emission line at 6.7 keV (the most important
emission line !) (Gilfanov, Sunyaev, Churazov
1987)
No longer optically thin, the surface brightness
is distort and you underestimate abundances in
the core of the cluster. You can measure it by
making the ratio of a supposed optically thick
(Fe He ?) and optically thin (Fe He ? at 7.90
keV) lines and see if there are deviation for the
prediction of the plasma code.
4
RESONANT SCATTERING IN PERSEUS ?
If there is resonant scattering, you
underestimate abundance in the prominent line at
6.7 keV you cannot reproduce the line at 7.9 keV
and you see residuals in the fit
MECS spectrum of the core (inner 8?) of the
Perseus cluster (from Molendi et al, 1998)
5
RESONANT SCATTERING IN PERSEUS ?
Correcting the apparent abundances
Using only He ?
Assuming optically thin emission
Molendi et al. 1998
But the excess could be due also to overabundance
of Ni respect to solar ratios, according to the
experimental evidence of a central enhancement of
SNIa ejecta (and Ni is produced only by SNIa) in
cD clusters (Dupke Arnaud 2001). The Ni He ? is
at 7.80 keV and you cannot distinguish it from
the Fe He ß at 7.90 keV with the MECS resolution
6
XMM OBSERVATION OF PERSEUS
MOS 1 image of the core of Perseus
PN light curve 10-13 keV
XMM has for the first time the combination of
resolution and effective area at high energies to
give an unambiguous answer. The Perseus cluster
was observed for 53 ks MOS and 25 ks PN but the
observation was badly affected by high
background. But we consider all the observation
exploiting the brightness of Perseus and
modelling the soft proton background. We
concentrate on hard bands (? 3 keV) which are the
one of interest to determine abundances of Fe and
Ni
7
OUR WORKING PROCEDURE ...
We model the soft protons wich contaminate the
spectra using in first approximation a power law
as a background model (the model is not convolved
via the effective area of the instrument)
8
IRON-K LINE TERMOMETER
9
OUR WORKING PROCEDURE ...
The temperature profile obtained by our best fit
model (mekal pow/b in the 3-13.5 keV) is in
good agreement with the temperature obtained by
the ratio of He ? to H ? (at 6.97 keV) Fe line.
It is also in agreement with the SAX-MECS
temperature profile.
10
RESONANT SCATTERING IN PERSEUS ?
EPIC-PN spectra 1-2 bin
Fe He ? 7.90 keV
Ni He ? 7.80 keV
As we can see leaving the abundance of Ni free,
the excess is due to this element and not to an
anomalosly high Fe He ? line. There is no
resonant scattering .
If you fit the XMM spectra with a MEKAL model you
find again an excess
11
ABUNDANCE GRADIENTS IN Fe AND Ni
The Fe and Ni abundance profiles obtained by our
best fit model (mekal pow/b in the 3-13.5 keV)
show a gradient. They are also in good agreement
with SAX-MECS determination, when we leave the
hypotesis of resonant scattering. Further tests
with 2T models in the broad band 0.5-10 keV (the
plasma is not consistent with being isothermal)
give similar results.
12
IMPLICATIONS OF NO RS

• The absence of a clear evidence of resonant
  scattering strongly points towards the presence
  of significant turbolence (Gilfanov et al. 1987
  Mathews et al. 2001)

AGN activity ? (caution see Sakelliou et al.
2002)

• It is a paradox that the first clear example of
  resonant scattering, NGC 4636 (Xu et al. 2002) is
  the most striking example of chaotic gas
  kinematics in the optical (Caon et al. 2000)

13
SUMMARY

• There is no need to invoke resonant scattering in
  the Fe He ? line in the Perseus cluster core ?
  the Fe abundance determination with optically
  thin emission models is reliable. It is now
  spectroscopically clear that the excess in the 8
  keV line complex is due to Ni. Fe and Ni show a
  gradient, adding another piece of evidence to the
  scenario of a central enhancement of SNIa ejecta
  in cD clusters.
• Turbulence could be an explanation for the
  absence of resonant scattering in the best
  candidates as M87 and the core of Perseus, maybe
  induced by AGN activity.

14
2T STRUCTURE
In the entire energy band 0.5-10 keV an
isothermal model is not a good fit and the data
requires at least two temperatures. For example
in the 0.5-1 bin we have
kTh EMh kTc EMc
?2/d.o.f. PNS (1Tpow/b
3-13.5 keV) 3.99 4.99 845/809
PNS (1T 0.5-10 keV) 3.29 5.75
2037/1299 PNS (2T 0.5-10 keV)
5.00 3.07 2.03 2.90 1734/1297
We simulate a 2T plasma as the best fit and we
find a temperature of 4 keV from the continuum in
the 3-10 keV band but of 4.7 keV from the Fe line
ratio.
15
COMPARISON WITH CHANDRA