Studying of multi-shell gas-puff Z-pinches using x-ray spectroscopy with spatial resolution

The gas-puff Z-pinch has been used for more than 50 years as a load on high-current generators. The study of this load was started almost simultaneously with the study of exploding wire loads and developed in parallel, mainly as a source of soft x-rays and neutrons.^1–20 During 2000–2010, against the background of success in the use of wire arrays, interest in gas-puff loads decreased. However, in recent years, new ideas on the use of the radiation from gas-puff Z-pinch loads have appeared, and during this time, the configuration of the loads has become more complicated. In order to study them, new diagnostics have been developed.^10,21–23 Currently, complex gas-puffs consisting of two or three gas shells are used, sometimes supplemented with a single wire, a wire array, or even an X-pinch.^{10–15,19,20}

The gas-puff Z-pinch load has a number of advantages over other cylindrical loads, such as cylindrical wire arrays and liners. One advantage is that the initial density distribution is axisymmetric. Gas-puff experiments can be carried out on relatively small generators without breaking the vacuum between pulses, which can lead to an increase in the number of shots per day. The nozzles for gas puffs can be designed to create unique profiles of distributed density, and the three plena in a triple coaxial valve make it possible to use up to three main gases in one shot (not to mention the possibility of additives to the gas shells to increase the diagnostic flexibility of a given gas-puff load). This feature was especially important in the development of radiation sources. The main disadvantage of gas puffs is that there is a limited number of gases that can be used. However, Ne, Ar, Kr, and Xe are very good for x-ray spectroscopy, since the Ne, He, and H-like lines of these gases cover a large energy range and allow the use of x-ray spectroscopy on generators with a wide range of output parameters. Experiments carried out on the 1 MA COBRA generator at Cornell University^21,23 have included investigations of the factors affecting the x-ray output of the K- and L-shells of the gases used, of the dynamics of the implosion and pinching, including the influence of a longitudinal magnetic field on the dynamics, and of the possibility of gas-puff based plasma opening switch creation.^{10,12–15,20}

Despite the creation of new diagnostics, x-ray spectroscopy continues to be the only diagnostic that provides information about the intensity, spectral, and ionic composition of x-ray radiation.^{2,6,7,10–12,20} X-ray spectroscopy with spatial resolution also provides information about the species involved in the compression of gas shells and the spatial distribution of various species, i.e., the dynamics of their pinching. In this paper, we show the benefits of the application of x-ray spectroscopy with spatial resolution for the diagnosis of multi-shell gas-puff Z-pinch loads. Data on the dynamics and plasma parameters of multi-shell gas-puff Z-pinches produced by the triple-nozzle system on the 1 MA generator will also be presented.

The gas-puff Z-pinch experiments were carried out over several separate sessions on the COBRA generator at Cornell University. COBRA produces a 0.9–1 MA peak current and a pulse duration of about 200–240 ns in its longer pulse mode of operation.^21,23 In the experiments discussed here, a triple-nozzle gas valve delivers two gas shells (an outer shell and an inner shell) and a central jet. The gap from the nozzle cathode to the anode ring was 25 mm. At the cathode exit plane, the inner and outer diameters of the nozzles are 6.2 and 4.2 cm for the outer nozzle and 3.6 and 1.4 cm for the inner nozzle, respectively. The opening diameter of the center jet nozzle is 1 cm. The radial density distribution peaked on the axis (Fig. 1).^1,2 In these experiments, Ne, Ar, and Kr of different pressures were used. In the experiments with three shells gas-puff load, the total mass line density was approximately 25–35 μg/cm.^14,24 A gas-puff pre-ionizing system to lower the initial load impedance and reduce unacceptably high load voltages and power reflection was implemented in the experiments.¹⁰ 

We have collected a huge set of spectra with different loads; conclusions seemingly drawn from one example spectrum shown in this paper are confirmed by many similar spectra.

The COBRA generator is equipped with a large suite of photon-sensitive diagnostics, including detectors covering several energy ranges from optical to x-ray.^{10–15,21–25} In this paper, we will focus only on the soft x-ray diagnostics used for gas-puff Z-pinches. The hardware is depicted schematically in Fig. 2(a) with a typical current waveform and x-ray signals shown in Fig. 2(b).

The main diagnostic device was a wideband focusing spectrograph with two spherically bent mica crystals with radii of curvature of 186 mm (FSSR-2, Focusing Spectrograph with Spatial Resolution). A detailed description of the spectrograph is given in Ref. 22. Notice that on one side of the double-crystal arrangement [orange and pink lines in Fig. 2(b), device 3], the distance from the crystal to the source is smaller, and the distance from the crystal to the film is larger, and vice versa on the other side. As a result, the magnification is slightly different across the entire spectrum and was corrected in analysis procedure that follows.

Another spherically bent mica crystal with a radius of 100 mm (position 5 in Fig. 2) was tuned to an angle of incidence of about 81° to obtain a quasi-monochromatic image of the pinch in the resonance line of He-like Ar. This image allows us to determine the size of the radiation source in the He-like Ar line in the direction of crystal dispersion.

For a complex load consisting of different materials, mica crystals that reflect in many orders of reflection can be more useful than crystals reflecting mostly in the first order. Using a mica crystal spectrograph FSSR-2, it is possible to record, for example, He- and H-like Ar lines in the third and fourth orders of reflection, Ne-like Kr lines in the second order, and He- and H-like Ne lines in the first order. Thus, in one shot, states of all ions in the gases can be recorded.

A third spectrograph shown in Fig. 2 had a spherically bent quartz crystal of very good quality with a radius of 180 mm (FSSR-1) and a crystal orientation of 1011 (2d = 6.66 Å). Since the quartz crystal reflects x-rays mainly in the first order, it was used to record only He- and H-like Ar lines. This helps in the accurate recognition of lines when using different gases in shells or wires instead of a jet. Good crystal quality helps to study individual spectral lines with good spatial and spectral resolution.

The optical quality of the mica crystals in FSSR-2 was not ideal, but it was good enough to focus the radiation in a line of ∼100 μm width corresponding to a spatial resolution along the radiation source in the axial direction of 0.2–0.5 mm at a distance of about 800 mm from the crystal. The quality of the quartz provided a spatial resolution of ∼50–100 μm.

All three spectrographs were located in a separate chamber. The entrance aperture to the camera with spectrographs is shown in Fig. 2(7). This arrangement made it possible to significantly reduce the level of background radiation produced by the COBRA generator. This is especially important when using image plates, the sensitivity of which significantly increases with increasing energy of the radiation. The energy of the background radiation, the source of which is the bremsstrahlung radiation of accelerated electrons, is much higher than the pinch radiation. Fuji BAS-TR image plates without a protective layer were used in the experiments.²⁶ For He- and H- like Ne lines, the use of these plates and very thin protective filters were very important. The total thickness of the protective filters in the experiments was 8 μm of mylar [four 2 μm thickness aluminized (250 Å) layers], and the absorption of which for the lines He- and H-like Ar and Ne-like Kr is negligible, but for He- and H-like Ne, it is about 90%–95% (Fig. 3).

Additionally, pinhole cameras with holes from 50 to 250 μm without a filter and with filters with different cutoff energies were used in the experiments. Calibrated PCDs with different filters were also used to estimate the soft x-ray output power and energy. Note that when comparing the spectra of the pinches recorded in different shots, the total radiation energy of the pinches recorded by the PCD with the Be filter (E >1.2 keV) was approximately the same and close to the maximum pinch energy recorded by the PCD. This indicates a good pinching and the reliability of comparing the intensity of the spectra.

Over several years of using x-ray spectroscopy in gas-puff experiments, a large set of data has been collected that cannot be presented in a single article. We will focus only on the main results, which have made a great contribution to experiments conducted with different goals.

The experiments used spectrographs with spherical mica and quartz crystals that made it possible to reliably separate the He- and H-like Ar and Ne lines although this was difficult with the mica crystal since the spectra of Ar in the third order and Ne in the first order were superimposed on each other. Figures 4 and 5 show the spectra recorded in shots with gas-puff loads consisting of Ar and Ne in different combinations (jet/inner shell/outer shell/). Figures 4(a) and 4(b) show that the presence of neon in the outer shell was not recorded in the spectrum. Both crystals showed the presence of only He- and H-like Ar lines. When using neon in the inner shell, on the FSSR-2 with a mica crystal, a spectrum of neon with significant intensity He- and H-like lines was recorded despite the absorption in the filters [Figs. 3 and 5(b)].

Figures 6(b) and 6(c) show pinhole-images of Ar–Ne pinch with cutoff energies that most closely correspond to the radiation energy of the He- and H-like lines of Ne and Ar. When comparing the image of the pinch in the He_α Ar line with the images recorded using pinhole cameras, it is clear that the monochromatic images coincide with the hottest areas of the pinch [Figs. 4(c), 5(c), 6(a), and 6(b)]. Whereas in almost all loads using neon in the inner shell, the radiation is seen along the entire length of the diode, as shown by the neon line radiation in the spectra and the corresponding pinhole-images [Fig. 6(c)]. When Ne was used in the central jet, no He- and H-like Ne lines were detected. Some of the radiation would have been absorbed in the outer layers of the plasma, and the rest in the filters. Other possibilities, such as a small number of ions in the jet or its complete ionization, seem less likely to us. First, the amount of Ne in the jet increased from shot to shot, but no He- or H-like Ne lines were detected. Second, before complete ionization, the Ne had to emit He- and H-like lines, which were not observed. Fully ionized Ne would require significantly higher temperatures that are not achievable under our conditions (see Fig. 10).

Figure 7 shows that the intensities of the Ar lines change mainly from one region to another, while the intensities of the neon lines change little. Changes in the intensities of the Ar and Ne lines do not correlate in space. This is especially clearly seen in the second spectrum profile, where the intensity of the neon lines is the maximum, and the intensity of the argon lines is the minimum. It follows that the He- and H-like Ne and Ar lines are most likely radiated at different times and possibly in different layers of the pinching plasma.

In addition to the Ar and Ne loads, experiments used Ar and Kr loads in various configurations; loads consisting of one of these gases were also tested. We will show below that the conclusions drawn from loads of relatively light gases, such as neon, cannot be carried over to other loads. Figure 8 shows the spectra recorded in the shot carried out with the heavy gas (Kr) in the inner shell (shot 3284) and in the jet (shot 3285) with Ar in the other nozzles. Here, we see that the intensity of the Ne-like Kr lines with Kr in the inner shell is significantly lower than in the shot with Kr in the jet. At the same time, the intensity of the He-like Ar lines is also higher when it is used in a jet. The total energy of the pinch radiation recorded by the PCD with Be filter (E >1.2 keV) is approximately the same in the two shots, which indicates good pinching and the validity of comparing the intensity of the spectra.

When the entire load consists of the same gas, it is very difficult to study the dynamics of the shells in the process of their pinching. However, adding small amount of a different gas, with a similar Z to the main one and radiating in the same spectral range, can greatly facilitate the task. As an example, Fig. 9 shows the spectral profiles recorded in shots with a load of Ar in all shells but with a small admixture of 4% SO₂ gas in the inner shell and the center jet.

A small amount of sulfur is sufficient to record intense He- and H-like S lines, demonstrating the high efficiency of the FSSR spectrograph.¹⁰ The He- and H-like S lines were recorded only when SO₂ was present in the jet. With SO₂ in the inner shell, S lines were not recorded. From the presented results, it can be concluded that in the case of heavier gases, such as Ar and Kr, the most active pinching occurs in the jet, while the shells serve to transfer their kinetic energy to the jet. In previous studies, it was shown that the intensity of x-ray radiation in loads with two inner shells is lower than in loads with three shells because of lower kinetic energy.¹⁰ It is likely that the kinetic energy of the outer shell is not the only mechanism for heating the load. For example, in Ref. 27, it was shown that there is a significant amount of energy coupled to the radiation field at the time of the pinch by the inner shell implosion. In Refs. 28 and 29, triple nozzle gas-puff implosions were described in terms of a “pusher” outer region that carries the current, a “stabilizer” inner region that mitigates the Rayleigh–Taylor instability, and a high density “radiator” central jet region that is heated and compressed to radiate. Additional studies are required to confirm any scenario of gas load dynamics with three shells. The FSSR spectrograph is a promising tool for a detailed study of the dynamics of heating and compression of a gas-puff load.

To analyze the results obtained, the spectra were simulated using the PrismSPECT program,^30,31 both separately for each gas and for a mixture of gases in the proportions: Ar/Ne: 90%/10%, 75%/25%, 50%/50%, and 25%/75%. The calculations include opacity effects under the assumption of a planar plasma geometry with a thickness of 1 mm obtained from these experiments. Some of the results of calculations of a mixture of neon and argon in the ratio of 1:1 are shown in Figs. 10–12. Calculations were performed for a wide range of temperature and density of the pinch plasma (Fig. 10). Figures 11 and 12 show the dependence of the intensity of the spectral lines on the temperature and density of the plasma.

The spectra calculated for a mixture of gases in the ratio of 1:1 and plasma layer thickness of 1 mm were closest to the experimental spectra. The experimental size of the pinch was determined from its image in the He_α Ar line, which also gave approximately 1 mm in radial direction. It should be noted that an accurate calculation of such a complex load is very difficult. Since the experimental spectra are time-integrated, it is possible that the gases participated in the pinching in layers and as a function of axial position; if they were mixed, the proportion of each in the radiation source could vary along the length, with radius and over time.

In the experimental spectrum, a very intense H_α –Ne line was recorded along with significantly less intense He_β and He_γ Ne lines. The resonance and intercombination lines of He-like Ar along with the satellites were also recorded in the pinch with Ar/Ne/Ar load. The H_α Ar line was recorded in two hot areas with size about 1 mm each in pinch height (Fig. 7). The calculations have shown that the He_β and He_γ Ne lines exist at a temperature near 100 eV, and their intensity increases with increasing density (Fig. 12). Above 250 eV, the intensity of these lines is insignificant. At the same time, the H_α Ne line, which also appears at temperatures of about 100 eV, has a maximum intensity at 250 eV, and then its intensity monotonically decreases but remains high (Fig. 12). The He- and H-like Ar lines have a low but noticeable intensity at electron temperatures above 300 and 500 eV, respectively, and ion density above 5 × 10¹⁹ cm⁻³ (Figs. 11 and 12). The intensities of all recorded lines increase with increasing plasma density (Fig. 11). The plasma parameters at the time of emission of the He- and H-like Ar and He-like Ne lines are very different.

Table I shows the plasma parameters estimated by fitting calculated spectra to the experimental ones shown in Fig. 7(a). When calculating the parameters of the pinch plasmas, almost 90%–95% absorption in the filters for Ne and 5%–10% for Ar lines were taken into account. The plasma parameters are given with a wide range, focusing on the existence of He- and H-like Ne and He- and H-like Ar lines in the spectrum and taking into account that these are time integrated spectra averaged in the axial direction within a radius of 2 mm. Note that the plasma parameters in the load of three shells (Ar–Ne–Ar) are higher than the parameters recorded in the load of the jet and shell (Ar–Ne or Ne–Ar).¹⁰ 

Summing up the data obtained in experiments with different gas loads, on a generator with a current of ∼1 MA and a current pulse rise time of ∼200 ns, the following conclusions can be drawn:

1. Spectroscopic studies have shown that when using a nozzle with two gas shells and a central jet, the dynamics of pinching of light (Ne) and heavy gases (Ar, Kr) is different. Neon forms dense and hot areas along the entire length of the high-voltage load region of the pulsed power machine only when used in its inner shell. Argon and Kr form such hot regions, too, but they are much smaller when they are used in the jet and the inner shell. It can be concluded that the outer shell mainly serves to transfer kinetic energy to the inner one.

2. When using shells of different gases, it was shown that the plasma parameters are very different for the various gases at the same point in the gas-puff load space. From this, we can conclude that light and heavy gases pinch at different times.

3. If the entire load consists of the same gas, adding small amount of a dopant gas that is not much different in Z from the main one and that emits in the same spectral range, can greatly facilitate the study of such a load. Due to the high luminosity of the FSSR spectrograph, the amount of a dopant gas can be very small and does not affect the dynamics of the pinch.

4. The experiments also made it possible to study the relative intensity of the Ne and Ar lines and their spectral position relative to each other. These studies have shown that when using mica crystals, the He_α Ar, He_β, and H_α Ne lines can be used for spectral studies with time resolution.

5. The above physical conclusions are only valid for generators with similar output parameters and nozzle design. The conclusion that is suitable for all gas-puff experiments is that x-ray spectroscopy with spatial resolution is a very effective diagnostic for large and complex gas loads and should be used in experiments as a routine diagnostic. X-ray spectroscopy also makes it possible to study the mechanism of heating the load material with a gas puff plasma source.

This research was supported by the National Nuclear Security Administration Stewardship Sciences Academic Programs through the Department of Energy under Cooperative Agreement No. DE-NA0003764.

The authors have no conflicts to disclose.

Tatiana Shelkovenko: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Validation (equal); Writing – original draft (equal). Sergey Pikuz: Conceptualization (equal); Data curation (equal); Investigation (equal); Methodology (equal); Writing – review & editing (equal). Niansheng Qi: Investigation (equal); Resources (equal); Writing – review & editing (equal). David Hammer: Conceptualization (equal); Funding acquisition (equal); Project administration (equal); Resources (equal); Supervision (equal).

The data that support the findings of this study are available within the article.