Benchmark Experiment for Photoionized Plasma Emission from Accretion-Powered X-Ray Sources

The interpretation of x-ray spectra emerging from x-ray binaries and active galactic nuclei accreted plasmas relies on complex physical models for radiation generation and transport in photoionized plasmas. These models have not been sufficiently experimentally validated. We have developed a highly reproducible benchmark experiment to study spectrum formation from a photoionized silicon plasma in a regime comparable to astrophysical plasmas. Ionization predictions are higher than inferred from measured absorption spectra. Self-emission measured at adjustable column densities tests radiation transport effects, demonstrating that the resonant Auger destruction assumption used to interpret black hole accretion spectra is inaccurate.

Figure 1

(a) Top-view schematic: the silicon sample is heated by x ray and expands to low density. X-ray diagnostics characterize the absolute x-ray irradiance on the sample. (b) Side-view photograph showing the $Z$-pinch radiator and the sample locations. (c) Measured spectral irradiance (black). Overlaid, typical AGN power law spectra with the same Si ionizing flux as the experiment with 1.3 (red) and 2.3 (blue) exponents. (d) The CH-tamped silicon is mounted edge-free enabling self-emission collection, with a 3 mm tall aperture to avoid plasma edges [Fig. 1]. The CH extends 1 mm beyond the Si length $L=3$, 6, and 12 mm, preventing lateral expansion. (e) Absorption spectral image (linear scale). (f) Emission image collected at 90° from the absorption line of sight (log scale).

Figure 2

Measured $1s\text{−}2p$ $K\alpha$ silicon plasma transmission spectrum. Top: Mean over 16 measurements with the standard deviation shown as grey background. Lines arising from Li-like (${\mathrm{Si}}^{+11}$) $1s^{2}2s$ ground state (G.S.) and $1s^{2}2p$ singly excited state (E.S.) used to infer electron temperature are indicated. Bottom: Absolute standard deviation in the mean transmission, with spectral mean (dashed).

Figure 3

Transmission compared with atomic and xstar models. Top: Models predict an overionized transmission at experimentally inferred conditions. Bottom: Tested models give reasonable agreement when $\overline{Z}\sim \mathrm{10}$, with adjusted conditions of $T_e=\mathrm{25}\mathrm{eV}$ and $2\times$ density, $n_e=1\mathrm{.}7\times {\mathrm{10}}^{19}{\mathrm{cm}}^{-3}$, for atomic, and with $4\times$ density, i.e., $T_e=\mathrm{33}\mathrm{eV}$, $n_e=3\mathrm{.}4\times {\mathrm{10}}^{19}{\mathrm{cm}}^{-3}$ for xstar. On each panel the charge state distributions for each calculation are shown in the inset.

Figure 4

Silicon emission collected for three column densities (contributing ions are labelled in top panel). Intensity units are proportional to $J.{\AA }^{-1}.{\mathrm{sr}}^{-1}.{\mathrm{cm}}^{-2}$. The observed column density is adjusted with the variable length $L=3\text{mm}$ (a), 6 mm (b) and 12 mm (c). The standard deviation $\sigma$ (green curves) and resolving power are given in each plot. The enlarged plot segments (insets) further illustrate the reproducibility.