Title: Iron K Spectra from L-Shell Ions
1Iron K Spectra from L-Shell Ions in Photoionized
Plasmas Work in Progress Duane Liedahl Physics
and Advanced Technologies Lawrence Livermore
National Laboratory
Motivations Si Ka fluorescence from L-shell ions
in HMXBs spectroscopic diagnostic
potential high spectral resolution at Fe K
provided by Astro-E 2 Focus resonant Auger
destruction Applications stellar winds in
HMXBs accretion disks in AGN and X-ray binaries
2ASCA data from Vela X-1 motivated a wind model
based on spherically symmetric mass loss
contours of log ionization parameter xL/nr2
Si charge state distribution
See Sako et al., ApJ, 1999
Model accounts for emission from H-like and
He-like ions of several elements Does not
account for fluorescence lines that were observed
3From wind model we can predict high-ionization
component of high-resolution Chandra spectrum
4Vela X-1data shows entire Si L-shell Ka spectrum
Several charge states are separated, although
each ionic component is a blend
Chandra spectrum
Mg-F
B
C
O
N
Be
5Same features are observed in other X-ray binaries
6Broad range of silicon charge states suggests
that fluorescence from iron L-shell ions should
be observed
7Iron K lines are used to probe black hole
accretion disks
8Models used to fit relativistic Fe K line involve
6.4 keV near-neutral line or H-like and
He-like lines no Ka from Fe L
Where are the K lines from Fe L-shell ions? It
was suggested that resonant Auger destruction is
responsible. (Ross, Fabian, Brandt 1996, MNRAS
see also Band et al. 1990, ApJ) How does silicon
observed in HMXBs respond to this process?
9Resonant Auger destruction is simply line
scattering with a high destruction probability
per scatter
mechanism operates for F-like to Li-like
ions (for Ka need vacancy in n2 shell)
Since destruction probability is high, a good
approximation is to zero out all K lines from Fe
L-shell ions right?
10We use an escape probability method to model
resonant Auger destruction
11Calculation of line optical depths requires
knowledge of level population distribution
appropriate to local plasma conditions
line optical depth depends on fractional
population of lower level
12Many iron Ka transitions terminate on excited
levels Example Be-like Fe XXIII
Calculations performed with the HULLAC atomic
physics package
13Fe XXIII illustrates the selective action of
resonant Auger destruction
This line terminates on ground
four lines to levels 6, 7, 8
models folded through DEfwhm 6 eV gaussian
resolution kernel
14Define an effective fluorescent yield to account
both for atomic physics and resonant Auger
destruction
15A better assessment of resonant Auger destruction
accounts for level population distributions
We have demonstrated the effect of Fe RAD for the
nebular case, that is, only ground levels are
significantly populated Should be adequate for
most HMXB environments, not so for disks
16Consistent treatment requires population kinetics
model for pre-ionization charge state (cf.,
Jacobs et al. 1989, Phys Rev A)
17Photoionization out of excited levels provides
access to different autoionizing levels,
resulting in a different K spectrum
high yield lines
line fluorescence yield
low yield lines
18We include a bright EUV radiation field to drive
the level populations
schematic of ion layer in an accretion disk
atmosphere
level populations for 9 lowest Be-like
levels for kT 80 eV Planckian
19Level population distribution can make a big
difference in both fluorescence line spectra and
ion fluorescence yield
Comparison of Fe XXIV spectra in zero-D
showing effect of different level population
distributions
20Resonant Auger destruction modifies the outgoing
spectrum but does not entirely quench emission
21Disk environment leads to enhancement of
effective fluorescent yield for this ion even
with resonant Auger destruction
22Li-like iron is not an exception similar
results are found for three other charge states
23Vertical disk structure calculations show that
column densities of Fe L ions are each a few
times 1018 cm-2 (see also Nayakshin
Kallman 2001, ApJ)
Calculation by Mario Jimenez-Garate figure
provided by Chris Mauche
24Ka emission from Fe L boosts theoretical Fe
line equivalent widths for expected accretion
disk parameters
ionic equivalent widths Li-like through F-like Fe
summed equivalent width
suggested range of column density found in
accretion disk atmosphere
25AGN models predicting relativistic O VIII
emission should include Fe L Ka calculations for
consistency
26Summary and Comments
Focus of this work is on assessing spectroscopic
effects of resonant Auger destruction Conclude
that it is not valid to turn off Ka lines from
L-shell ions when modeling disks That L-shell
ions are not required in fits to AGN spectra
remains puzzling but will be better tested with
Astro-E 2 Fe K lines from L-shell ions should be
observed in some HMXBs Connection between HMXBs
and AGN in this context? HMXB spectra can be
used to exercise spectral models by
providing constraints based on observations
feed back into disk models Resonant Auger
destruction can be turned into a new class of
plasma diagnostic