Production of photoionized plasmas in the laboratory with x-ray line radiation

In this paper we report the experimental implementation of a theoretically proposed technique for creating a photoionized plasma in the laboratory using x-ray line radiation. Using a Sn laser plasma to irradiate an Ar gas target, the photoionization parameter, $\xi =4\pi F/N_e$, reached values of order $50\text{erg}\text{cm}{\text{s}}^{-1}$, where $F$ is the radiation flux in $\text{erg}\text{c}{\mathrm{m}}^{-2}{\text{s}}^{-1}$. The significance of this is that this technique allows us to mimic effective spectral radiation temperatures in excess of 1 keV. We show that our plasma starts to be collisionally dominated before the peak of the x-ray drive. However, the technique is extendable to higher-energy laser systems to create plasmas with parameters relevant to benchmarking codes used to model astrophysical objects.

Figure 1

Inset: Schematic of the gas cell. Main: Experimental schematic showing the drive laser beams incident upon a thin layer of Sn coated on top of CH. The x-ray target was placed over a 0.5 mm opening made in a reentrant plug which was fitted inside the “east” window of the gas cell. This allowed observation of the K-shell florescence close to the x-ray source by a spherically curved mica crystal spectrometer through a Kapton window mounted on the “south” side of the gas cell. The dispersion plane was vertical, allowing us to spectrally resolve and to spatially resolve x-ray emission along a direction normal to the laser focal plane. Two flat silicon crystal x-ray spectrometers were used to spectrally characterize the x-ray target, with one viewing the laser-irradiated side of the target and the other through the target and a second Kapton window. In addition, through this window a streak camera coupled to a flat HOPG crystal measured the temporal profile.

Figure 2

(a) Raw spectrum for a typical shot taken with a flat Si crystal spectrometer. (b) Lineout from the raw image. The energy calibration of the spectrometer was performed using a test spectrum from a Ca target with well-known He-like and Li-like lines. (c) Lineout combining several shots with various Bragg angles, and scaled according to laser energy, to assemble the whole L-band spectrum from Sn.

Figure 3

(a) Lineout of the streaked x-ray emission for Sn L-band radiation measured with $\sim 30$ ps resolution. (b) Raw image of the x-ray streak data for the line group from $\sim 3700–3850\mathrm{eV}$ seen in Fig. 2. The lineout in (a) shows the timescale. (c) Raw x-ray pinhole image of the Sn plasma emission in the keV range and a vertical lineout. The FWHM of the x-ray emitting region is $\sim 300\mu \mathrm{m}$ in this case.

Figure 4

(a) Raw CCD image for a 500 mbar Ar gas fill (678 J laser energy). A shadow region is evident between those where the Ar is viewed and where plasma from the expanding laser-driven material is visible as it moves away from the gas cell window. Hard x rays from the Sn laser plasma cause speckle across the chip. In (b) the shot at the same gas pressure (787 J laser energy) has the 218 nm Sn layer replaced with 186 nm Al, and there is a dramatic drop in the emission of Ar $\mathrm{K}\alpha$ and virtual disappearance of $\mathrm{K}\beta$ emission. The calibration scale is the same for both images. (c) Raw spectra at varying gas pressure, taken $\sim 0.5–0.7\mathrm{mm}$ from the x-ray drive source. For 1000 and 300 mbar there is a similar level of maximum ion stage observed, but as the pressure drops to 152 mbar and then 50 mbar the maximum ion stage also decreases. Red squares mark the expected positions of the Ar $\mathrm{K}{\alpha }_1$ lines from different ion stages, where the charge denoted refers to that of the initial ion prior to creation of the inner-shell vacancy by photoionization.

Figure 5

(a) Lineout from two shots with a 50 mbar gas fill close to the source at $z=0.5\text{–}0.7$ mm with incident flux $\sim 10^{19}\left(\mathrm{erg}/\mathrm{c}{\mathrm{m}}^2\right)/\mathrm{s}$. The images are despeckled before the lineouts are taken (pixels replaced with median of a surrounding $3\times 3$ grid, conserving overall signal). For this lowest pressure case shot-to-shot variation in the profile can be seen more strongly than at higher pressures due to lower signal levels. (b) Raw data lineout (0.5–0.7 mm) from a shot with 152 mbar gas fill and a similar incident flux to (a). See the text for discussion of the simulations. In both figures we mark the positions of the expected $\mathrm{K}{\alpha }_1$ lines from different ion stages.

Figure 6

(a) Geometry of the recessed target. The flux profile at the focal plane is taken from the pinhole data above. (b) Profile of the peak flux calculated as a function of the distance, defined as $r$ in panel (a) of the figure, and at a distance $z=0.6$ mm. Away from the center, the recess walls block flux and thus the flux is restricted in spatial extent and thus value. (c) Peak flux at $r=0$ calculated as a function of $z$. We can see there is a steep gradient in flux as discussed in the text.

Figure 7

(a) Calculated emission history using a simple plasma code as described in the text. The ion stages are those emitting the photons. The temperature rises to above 20 eV before radiative losses cool the plasma; the main emission is predicted to occur when the temperature is 15–20 eV. (b) Temporal history of the $\xi$ parameter for the lowest two pressures studied. The ratio of collisional ionization to photoionization for $\mathrm{A}{\mathrm{r}}^{3+}$ shows that at early times prior to the peak flux we have a photoionization-dominated plasma.

Figure 8

In this plot we show the time history of average ionization for three gas pressures. The incident flux of L-shell radiation is ${10}^{19}\left(\mathrm{erg}/\mathrm{c}{\mathrm{m}}^2\right)/\mathrm{s}$ and an additional 2% population of fast electrons is added assuming an effective temperature of 250 eV. This adds about 25% to the energy deposited. More detailed modeling would be needed for a full understanding but this simple calculation shows that the effect is more pronounced for higher gas pressures. This is consistent with the higher ionization observed for higher gas pressure in the data.