Transcript neon photoionization experiments on the Z-machine

X-ray Spectral Diagnostics of Neon Photoionization Experiments on the Z-Machine
David H.
1,2
Cohen ,
Joseph J.
2
MacFarlane ,
James E.
3
Bailey ,
Duane A.
4
Liedahl
(1) Swarthmore College, (2) Prism Computational Sciences, (3) Sandia National Laboratory, (4) Lawrence Livermore National Laboratory
Gas Cell Experiments
Introduction: We have begun a campaign of ride-along shots on the Z-machine at
Sandia in order to study the response of low-density gases to irradiance by strong xrays. Our interest is in the excitation/ionization kinematics and the physical conditions
of the x-ray photoionized plasma. We would like to minimize gradients, shocks, and
collisional effects and isolate the atomic physics. The motivation--described below--is
primarily to model astrophysical photoionized plasmas and benchmark codes used to
study astrophysical sources, but there are also basic physics applications, as well as
potential applications to gas filled hohlraums and x-ray laser sources.
The experimental package consists of a cm-scale neon-filled cell with mylar windows,
mounted several cm from the Z-machine’s current return can, inside of which is the
pinch itself -- an imploding tungsten wire array, that can generate upwards of 200 TW
of x-rays in a ~10 ns pulse, characterized by an emission temperature of roughly 200
eV.
Experimental Design
Accretion onto black holes or neutron stars can heat plasma to a
few 100 million K.
The circumstellar gas near the source of hard x-rays is highly ionized--overionized for its
temperature compared to coronal equilibrium--and produces an x-ray spectrum dominated by
radiative recombination continua and recombination cascades.
Hydrodynamic Simulations of the Gas Cell
Viewfactor simulation output is used as input to our
hydrodynamic simulations.
Conclusions from hydro: To optimize uniformity of the physical conditions, a
lower gas density is desirable. To minimize compression of the gas by the mylar
walls, thicker walls are desirable. By doping the walls or coating them with thin
foils, we may be able to optimally filter the x-rays flowing into the gas cell as well
as putting off the implosion of the gas by the wall plasma until very late times.
Spectral Diagnostics
Backlit Absorption
X-ray binaries (artist’s rendition, above left) generate hard X-rays as material flows off a
giant star and accretes onto a compact object. This hard X-ray emission is reprocessed in
the wind of the normal star, producing recombination radiation, which is the primary
component of the observed spectrum when the compact object is eclipsed by the normal
star (schematic, above middle). The conditions are traditionally described by the
ionization parameter (flux/density), which is a measure of the dominance of radiation
processes over collisional processes. Contours of constant ionization parameter
(logarithmic, in cgs units) are shown for the HMXRB Vela X-1 (above, right; Figure from
Sako et al. Ap.J., 525, 921 (1999)). Our previous experiments have achieved an ionization
parameter of ~7, but significantly higher values are possible.
The time-dependent radiation temperature incident on the front of the gas cell (left). On
the right, we show, for one simulation time, the incident spectrum (red) and blackbody
equivalent (black) for one point on the front of the cell. Note that incident spectrum is
quite non-Planckian.
Hydro Simulations of the Experiments
The time-integrated absorption spectrum from
our first round of shots (top, above) are wellmatched by our calculations (lower, above).
Ultimately, we plan on making simultaneous, time-resolved absorption and emission
spectroscopic measurements of the gas cell, which can be filled with a variety of gases,
including mixtures. The absorption spectroscopy uses the pinch itself as a backlighter.
Experiments already completed have used 30 Torr neon (nion ~ 1018 cm-3) observed in
absorption with a time-integrated spectrometer.
Calculation of iron emission spectrum for a photoionized plasma (left) and a collisional
plasma (right). Though the ionization distribution is the same in both models, the spectra
are quite different (and the photoionized plasma is much colder). (Figure taken from
Liedahl et al. Ap.J., 350, L37 (1990).)
Motivation
Many astrophysical x-ray sources--especially those powered by accretion onto a compact
object (black hole or neutron star)--are photoionized, as opposed to the collisional plasmas
present, for example, in the solar corona. Yet almost all of the x-ray spectral modeling
effort in the astrophysics community has focused on coronal plasmas.
But with the advent of high-resolution (l/Dl ~ 1000) x-ray spectroscopy in the last several
years, a wealth of spectroscopic data for photoionized sources has been generated, and the
level of detail and accuracy demanded of models has increased dramatically.
A series of Ka absorption spectra,
representing three different sets of
physical conditions in the gas cell
(warm to hot from top to bottom).
Spectral resolution is 5 eV (FWHM).
Experimental Design
VisRad Viewfactor Simulations of the X-ray Source
Viewfactor simulations of the pinch and target are used to calculate the
incident spectrum on the gas cell, and also to investigate the spatial
uniformity of the irradiance on the cell.
By producing and measuring a well-characterized x-ray photoionized plasma in the
laboratory, we hope to benchmark the codes used to model x-ray spectra of active galactic
nuclei, x-ray binaries, and other astrophysical sources for which photoionization dominates
collisional ionization.
The Chandra telescope (above left) and XMM telescope, both launched in 1999, are the first high-resolution
spectroscopic missions for the astronomical community. Both employ gratings (above center), X-ray
missions planned for the future include arrays of telescopes (Constellation-X, above right) and use
microcalorimeters for high spectral resolution.
Temperature (left) and density (right) at four different times in the simulation of our
experiments. The left side of the 1.1 cm cell is at x=0.0 cm. The radiation is incident
from the left. The initial neon gas density is 1018 cm-3. Note the shock waves launched
from the mylar walls at late times. Note also the radiation wave (as evidenced by the
temperature gradient) traversing the gas at t=100 ns.
A section of our raw data from the same
shot (above). Note the very high principle
quantum number lines of He-like neon.
Spectral Diagnostics
Emission
Same as the first simulations, but with mylar walls four times thicker (5.6 microns). Note
that the heating of the gas is reduced by the absorption in the thicker walls. And also note
that the shock waves are launched into the gas somewhat later than in the first simulation
(at t=100 ns, the shock has not progressed as far into the gas).
Two snapshots (an early time, during the low-temperature foot of the pulse on the left and a
late time from the high-temperature peak) from a VisRad simulation of the imploding pinch
and gas cell. The four cells, in each snapshot are a complete view of the pinch, current return
can, and target (represented by the flat square) (top left), a zoomed-in view of the pinch
through the current return can (the opaque slats of which are represented by ‘scaffolding’ in
the front so as not to obstruct the view in this representation) (top right), a view from inside
the can, looking through the slats at the target (lower left), and the instantaneous spectrum
incident on the front of the gas cell (lower right). (Figures courtesy of G. Rochau.)
A spectral simulation of neon, based on a representative hydro calculation. The
intrinsic spectrum (left) shows numerous lines of B-like through H-like neon, as well
as a recombination edge (near 1200 eV) and recombination and free-free continua. At
2 eV (center) and 5 eV (right) spectral resolution, many of the features are blended.
The 2 eV resolution spectrum is probably good enough to quantitatively test
atomic/spectral models, whereas a lot of information is lost at a resolution of 5 eV.
This simulation has the thicker mylar walls (5.6 microns) but only one-tenth the initial
neon density (1017 cm-3). The gas temperature is higher, since the gas is almost completely
optically thin to the x-rays. There is no longer a radiation wave diffusing through the gas.