1 Introduction

Demonstration of lasing towards shorter wavelength region of the electromagnetic spectrum, specially in soft X-ray region, has always been the centre of interest for the scientific community involved in the development of lasers. The motivation behind it lies in the fact that X-ray lasers have many potential applications in wide areas of science and technology, like nano-imaging including 3-dimensional molecular imaging, dense plasma diagnostics, lithography, holography, study of radiation-induced damage in bio-molecules [1,2,3,4,5,6,7]. In fact, if such sources are available in a compact table-top form in the laboratories, several new applications may come up in future. Pumping through fast capillary discharge is one of the promising techniques which has great potential for making table-top X-ray laser systems. First time, this technique was utilized to observe soft X-ray amplification in C VI Balmer-α transition at 18.2 nm driven by recombination pumping [8]. However, this was a kind of ablative discharge where the gain medium of carbon plasma was formed by the ablation of capillary wall made of carbon-containing material. Such a laser had a small gain-length product limited by the non-uniformity of the plasma column beyond few millimetres.

The journey of fast capillary discharge scheme towards X-ray laser was boosted further with the first experimental demonstration of soft X-ray lasing at 46.9 nm based on collisional excitation pumping in Ne-like Ar by Rocca et al. [9]. Here, the gain medium was formed by the discharge through gas-filled capillary and not by the ablative discharge. The axial uniformity could be obtained over a larger plasma length inside the capillary due to the introduction of a small current (few amperes) pre-pulse a few microseconds before the main pulse. This helped in getting larger gain-length product and as a result, the scheme progressed well with the development of soft X-ray laser working even in the saturation regime [10, 11]. This was further extended for making very compact X-ray laser system running at few hertz repetitive rate [12, 13]. Subsequently, various groups across the globe also reported successful demonstration of this X-ray lasing and contributed significantly in this area of research [14,15,16,17,18,19,20,21,22,23,24,25,26,27]. Apart from argon gas, the scheme was successfully utilized to demonstrate soft X-ray lasing in some other elements also like Ne-like S at 60.8 nm [28] and Ne-like Cl at 52.9 nm [29].

Efforts are presently going on worldwide to demonstrate lasing at shorter wavelength region using the pumping scheme driven by fast capillary discharge. The most promising candidate is the Balmer-α transition at 13.4 nm in H-like N (N6+) ion which is based on recombination pumping of a nitrogen-filled capillary by fast electric discharge. It requires generation of fully stripped nitrogen ions (N7+) by rapid ionization at pinch formation. During subsequent expansion of the axially pinched plasma, rapid cooling takes place which can lead to dominant 3-body recombination of N7+ ion into H-like ion, i.e. N6+. The feasibility of this lasing through capillary discharge was first proposed by Vrba et al. through 1-D MHD simulations [30]. These simulations predicted that a gain of 1.0 cm−1 can be achieved for Balmer-α transition of H-like N ions when a discharge current of ~50 kA with a quarter period (T1/4) of ~40 ns is passed through an alumina capillary of inner diameter 3 mm filled at initial gas pressure of 5 mbar. Here, the wall ablation of capillary was neglected. However, taking into consideration the wall ablation, the gain was estimated to be much smaller ~0.11 cm−1 [31]. Kolacek et al. [32] did the first experimental effort in this direction with a discharge current of ~56 kA having quarter period ~100 ns passing through nitrogen-filled alumina capillary of 3 mm inner diameter. Pressure in the capillary was optimized to record fast X-ray pulse (5–10 ns) in vacuum photo-diode as a signature of pinch emission. Although some line emissions from N5+ state were reported in their spectroscopic study, the most important emission at 2.8 nm (He-α) was not observed. Kampel et al. [33] utilized a pulse power generator delivering 60 kA peak current with ~70 ns quarter period to excite nitrogen Z-pinch plasma in a 90-mm-long and 5-mm-diameter quartz capillary. However, they could demonstrate only ~10% against the required 50% abundance of N7+ ions and concluded that higher power driver is needed to achieve this. Spectra generated from similar current profile were studied and analysed in detail by Gissis et al. [34] showing He-α and Ly-α emissions by N5+ and N6+ charge states, respectively. They pointed out that, in order to achieve higher ionization degree, further upgradation of the experimental apparatus is needed for driving yet higher discharge current. In order to get faster heating as well as faster recombination, triangular pulse shaping was also tried with ~50 kA peak current and ~50 ns pulse width, in a nitrogen-filled capillary of 3 mm diameter and 75 cm length [35]. However, the plasma density and the temperature produced in their experiment was found to be insufficient for the required X-ray lasing. Although the efforts have not been successful till date in realizing this X-ray lasing, all these reports strongly suggest that further experimental investigations are required with much higher pulse power driver. In recent experiments done in our laboratory on capillary discharge argon laser, it was found that faster current driver can relax the high peak current requirement by substantial amount for getting similar energy in the laser pulse [36]. However, it will be of great interest to investigate the nitrogen discharge plasma with a high as well as fast discharge current while looking for H-like N recombination X-ray laser at 13.4 nm.

In this paper, we report the experimental efforts that have been carried out to study recombination X-ray lasing in nitrogen utilizing our recently upgraded capillary discharge system capable of delivering peak discharge current more than 100 kA with a quarter period of ~46 ns. Alumina capillary of inner diameter 2.8 mm and length 96 mm was used in these experiments. We report the first temporal and spectroscopic study of nitrogen Z-pinch plasma at such a high and fast discharge current which was not reported in literature. We also report the first experimental evidence of the effect of pre-pulse on the spectrum of the highly ionized nitrogen Z-pinch plasma, confirming the important role of the pre-pulse in attaining higher charge states of nitrogen. Section 2 briefly describes our capillary discharge system used in these experiments. The results obtained in these studies are discussed in Sect. 3.

2 Experimental set-up

The experimental conditions in terms of fast and high current as well as high gas pressure required for the X-ray lasing in nitrogen, as predicted by simulations [26, 30], are very stringent and have been achieved in our laboratory. The fast capillary discharge system used for the experiments described here is a fully upgraded version of our older system described in Ref. [27]. A schematic diagram of the upgraded system is shown in Fig. 1. Here, the Marx-bank-based charging unit has been replaced by an in-house fabricated Tesla Transformer-based charging unit. It is utilized to charge a waterline capacitor (C ~ 6 nF) up to 400–500 kV in 3–4 μs, which is then rapidly discharged in a few tens of nanoseconds through a self-triggered spark gap switch connected in series with the nitrogen-filled capillary in order to generate a Z-pinch nitrogen plasma. The capillary used here was made of high-purity (99.7%) alumina, with inner diameter of 2.8 and 96 mm length. In order to make the current pulse faster, the length of the capillary was reduced from 150 mm (used earlier) to 96 mm. This lowered the inductance of the discharge circuit significantly making the current faster as also reported by Frolov et al. [37]. Although this has been done at the cost of reducing the length of the gain medium (plasma here), this reduced length of 96 mm should still be sufficient enough to generate a detectable XUV lasing pulse in case the lasing conditions for 13.4 nm are achieved. It should also be noted here that the length of the gain medium can be easily altered in this system for gain-length studies by changing the length of the high-voltage electrode going inside the capillary. This makes the set-up flexible enough for carrying out such studies in future. Also, this does not require changing the capillary length which otherwise may affect the current profile leading to erroneous result in the gain-length studies. The discharge geometry was modified wherever possible to minimize the inductance. There is an arrangement to pass a pre-pulse current (20–120 A) through the gas-filled capillary in order to pre-ionize the gas before passing the main discharge current. This pre-pulse is required to maintain uniformity over the entire length of capillary and minimize the growth of instabilities during the pinch formation. The requirement of higher density in a recombination-pumped X-ray laser demands higher nitrogen gas pressure (a few mbars) in the capillary. However, higher gas pressure in the capillary affects the vacuum on the diagnostic side where a detector-like micro-channel plate (MCP), which requires high vacuum conditions, is to be used. For achieving these conditions, a small orifice of 0.6 mm of diameter and made of copper–tungsten alloy was placed 3 mm away from the exit end of the capillary. This orifice separated the capillary from a small buffer chamber connected with vacuum systems. On the other side of the buffer chamber, there was a long metallic tube housing X-ray diagnostics like X-ray diodes or MCP. Another vacuum pump was connected close to the X-ray diagnostics to further improve the vacuum. This arrangement maintained the required vacuum level through differential pumping between the diagnostic side and the capillary. The alignment of orifice (0.6 mm) was very critical because it had a diameter comparable to that of the plasma column. Hence, the orifice was designed in such a way that it was guided by the outer diameter of the capillary, precisely sitting over it. This provided a self alignment of the orifice with respect to the exit end of the capillary. Of course, the design of the orifice also took care of the provision for gas entering into the capillary as well as its vacuum sealing with the surroundings. A quadrant vacuum diode was used to record the temporal profile of the capillary emission and also in locating the centre of the beam. The spectral information was obtained by using a free-standing transmission-grating spectrograph, and an MCP followed by CCD as detector. The voltage at the waterline capacitor was measured using a capacitive divider, whereas the discharge current through the capillary was recorded using a calibrated Rogowski coil. Typical waveforms of waterline voltage and discharge current are shown in Fig. 2. The discharge current has a peak value of ~105 kA. However, for safety reasons, the system was operated at lower values for the present experiments discussed here. Figure 2 also shows an X-ray laser pulse from argon discharge plasma (46.9 nm) under identical conditions.

Fig. 1
figure 1

A schematic diagram of the experimental set-up of the capillary discharge X-ray laser system

Full size image
Fig. 2
figure 2

Typical voltage and current waveforms with a peak current at 105 kA. The XUV laser pulse, obtained simultaneously at 46.9 nm from an argon plasma, is also shown here

Full size image

3 Results and discussion

Capillary discharge experiments were carried out in nitrogen gas filled in a ceramic capillary of length 96 mm and diameter 2.8 mm at discharge currents ranging from 60 to 95 kA. The quarter period of the discharge current in the present geometry was ~46 ns. The temporal profile of the emission from capillary was recorded by a quadrant vacuum diode and a fast oscilloscope (LeCroy, 1 GHz, 10 GS/s). A fast rise pulse with a rise time of 5–8 ns was detected along with Bremsstrahlung emission from plasma. The amplitude of this fast pulse was optimized by varying the gas pressure in the capillary from 2.5 to 11.0 mbar. This optimization is shown in Fig. 3a–c recorded at different gas pressures. In the temporal profile of the diode signal, the three coloured lines correspond to signals from the three quadrants. The small variation in the amplitude of these three quadrant signals may be attributed to small misalignment of the 0.6 mm pinhole with respect to the capillary axis. This will have effect on the individual quadrant signals in terms of slight difference in their amplitudes. The fast pulse recorded in each quadrant has a larger amplitude as compared to the background plasma emission for a narrow pressure range, as can be seen in Fig. 3. This enhanced fast pulse is seen in the gas pressure range ≈4–8 mbar for peak discharge current of 85 kA. It appears at the instant when there is an abrupt change in the slope of the discharge current waveform which is visible by carefully observing the current profile shown in Fig. 3. This abrupt change in the slope is caused by the significant increase in the inductance of the plasma column when it acquires a minimum diameter due to pinch formation [13]. This means that the fast pulse appears at the instant of pinch formation. This clearly signifies the fact that this fast pulse originates from the significant heating of plasma for a short duration at the time of pinch formation. The instant of pinch formation, which is reflected by the abrupt slope change in the current profile, also changes with the gas pressure and is clearly seen in the recorded profiles at different gas pressures shown in Fig. 3. Here, it is important to note that for an efficient discharge, the pinch formation should not occur much before the peak of the discharge current [38]. Furthermore, if the discharge current continues to increase after the plasma implosion, it will spoil the rapid cooling due to the expansion of plasma column and become unfavourable for recombination pumping. In view of these facts, the gas pressure has been increased beyond ~5 mbar to obtain the pinch formation after the current peak as shown in Fig. 3. Close to the current peak, enhanced amplitude of the fast rise pulse has been observed which falls down when the gas pressure is changed significantly on either way. However, in some conditions, specifically at lower gas pressures, steep rise of ~2 ns could be seen at the rising edge of this fast pulse as shown in Fig. 3d, which is also very encouraging for further experiments. The temporal response of the diode was sufficient enough to measure this steep rise, as it was earlier used to detect the 46.9 nm argon X-ray laser pulses of ~1 ns duration. Since the plasma is heated to higher temperature at lower gas pressure, the observation of steep rise could be a signature of plasma emission corresponding to short-lived higher charge states of nitrogen. It was found that the amplitude of the fast pulse is strongly dependent on the discharge current amplitude as well as the gas pressure.

Fig. 3
figure 3

Temporal profile of the quadrant diode signal along with the current profile obtained for different gas pressures: a 2.5 mbar, b 7.0 mbar, c 9.0 mbar and d 3.0 mbar (zoomed in time-scale), for a peak discharge current of ~85 kA

Full size image

In order to further investigate the spectrum of emission from capillary, the emitted radiation was dispersed by a free-standing transmission grating (700 lines/mm) and the time-integrated transmission spectrum was recorded on a micro-channel plate which was imaged on a CCD camera. For initial spectral measurement, the time-integrated mode was preferred over the time-resolved one as it was easier in this mode to look for the possible lasing (13.4 nm) signature whose intensity level as well as instant of occurrence is not known precisely in advance. Once its faint signature is identified, the measurement can be carried out into time-resolved mode. Also, time-integrated spectrum can provide in a single shot a larger view of the various emission lines generated at different times in the plasma. The experimental arrangement for recording the spectrum is shown in Fig. 4. Before placing the transmission grating, the axial position of the X-ray emission from capillary was precisely obtained with quadrant diode detector taking reference of the 46.9 nm argon X-ray laser beam generated from the capillary filled with argon. Figure 5 shows the recorded transmission grating spectrum at a discharge current of 95 kA and an optimized nitrogen gas pressure of 8.5 mbar. Only one side of the zeroth order is shown in the spectrum for better clarity. The lower part is the binned intensity profile of the spectrum showing various X-ray emission lines from nitrogen Z-pinch plasma. Zeroth order recorded in the transmission spectra was used as a reference for zero dispersion, and the distances involved in the spectral measurement set-up were used to calibrate the dispersion direction in terms of wavelength. Lines in the spectra have been identified using NIST atomic spectra database [39]. Careful identification of the spectral lines, especially towards the shorter wavelengths, reveals the nitrogen higher charge states. The spectrum was recorded initially at different distances between the grating and the MCP like 8.5, 18.5, 51 and 108, with different spectral resolutions. The best resolution was 0.2 nm achieved for 51 cm distance between the MCP and the grating. This was attributed to the observation that increasing the distance further (>51 cm) the beam divergence increases the width of the image of the grating-slit on the MCP to a level where the resolution starts becoming poorer.

Fig. 4
figure 4

Diagram of the experimental set-up for the spectroscopic study

Full size image
Fig. 5
figure 5

Transmission grating spectrum at 95 kA current and 8.5 mbar gas pressure, in presence of the pre-pulse

Full size image

In the spectrum, two prominent spectral lines, one at 2.8 nm and the other at 2.1 nm, were observed along with their various diffraction orders. Apart from that, there are several comparatively weaker spectral lines at higher wavelengths also, corresponding to different charge states of nitrogen. The gas pressure was optimized by looking at the intensity of the 2.8 and 2.1 nm lines. Whereas the emissions at 2.8 nm could be undoubtedly identified to be He-α emission from N5+ ion, the emission at 2.1 nm was quite puzzling. This emission matches with the position of Ly-β line of N6+ ion. However, there was no emission observed corresponding to the Ly-α line of N6+ ions at 2.4 nm wavelength. Given the fact that Ly-α line emission is always much stronger than Ly-β emission, it is quite difficult to assign the line emission at 2.1 nm to Ly-β line of N6+ ion, in the absence of the Ly-α line. There is another possibility that this line could be He-α emission from O6+ charge state of oxygen. Oxygen may come in the nitrogen plasma as impurity either from the gas itself or by the wall ablation of alumina (Al2O3) capillary. It was observed that the intensity of both lines, i.e. 2.8 and 2.1 nm varies with the nitrogen gas pressure in the capillary. Also, when the nitrogen gas was replaced with the argon gas and the same discharge current was used to excite the plasma, both these lines were found to disappear. These observations pointed out initially towards the contribution of oxygen impurity in the gas as the possible cause for 2.1 nm line emission. In order to further investigate it, experiments were again conducted with nitrogen gas of high purity 99.999%. There was no change seen in the intensity of 2.1 nm line. Further, experiments were conducted under similar conditions with air (~21% oxygen) and then pure oxygen gas in capillary with a view that if the source of 2.1 nm line is oxygen, it must get enhanced with the increase in oxygen content. Interestingly, there was no enhancement seen in the intensity of this line even with pure oxygen gas in capillary. This observation was quite surprising and puzzling as well. This actually points out that the origin of this line is not linked to the oxygen impurity in plasma coming from either the nitrogen gas or the wall ablation of alumina capillary. Considering the above observations, it appears that the origin of this line has to be related to nitrogen content and to be more specific, the only possibility lies for this line to be Ly-β of N6+. However, a definitive statement can be made in this regard only after extensive spectroscopic simulations of such hot and high-dense plasma including all the excitation, de-excitation, ionization and recombination processes, and considering the opacity effects.

The effect of current pre-pulse (20–120 A) on the emitted spectrum of nitrogen Z-pinch plasma was also studied. It was found that at higher gas pressures (>5.0 mbar), the spectral lines at 2.8 and 2.1 nm disappear if there is no pre-pulse. This is shown in Fig. 6. This indicates that pre-pulse helps in achieving higher charge state of nitrogen ions. However, no significant change was observed in the spectrum on increasing the pre-pulse amplitude from 20 to 120 A. The pre-pulse amplitude could not be increased beyond 120 A due to restriction on the pre-pulse electrode geometry in our capillary discharge system. However, a stronger pre-pulse will also be explored in near future. During the study, it was found that if the gas pressure is reduced below 3.0 mbar, the two emission lines (2.8 and 2.1 nm) re-appear even without the pre-pulse. The possible reason for this lies in the fact that plasma reaches a higher temperature at smaller gas pressure, if the driving electrical current remains the same and the temperature is sufficient enough to get nitrogen ions of higher charge states even when there is no pre-pulse. It may be noted here that there is no significant change in the amplitude or duration of the discharge current when the gas pressure was changed during the experiments. Another interesting observation is that when the gas pressure was reduced to below 3.0 mbar, a faint appearance of Ly-α of N6+ ion at 2.4 nm spectral line is seen lying between the lines 2.8 and 2.1 nm, as shown in Fig. 7. This implies that higher temperature of plasma, needed to get sufficient formation of N6+ charge state, is reached. An estimation of density and temperature will be of much relevance here. In order to get the electron density of the pinched plasma, experiments of X-ray pinhole imaging of the plasma column through the orifice need to be carried out which has not been conducted as yet and is planned to be done in near future. However, an estimate of the density can be obtained from the compression ratio of the plasma taking to be approximately 1:100 as is observed in various reports [33, 34, 40]. It is to be noted here that the discharge current amplitude in our system is higher and hence the compression ratio may be higher than what is estimated here. For a filling gas pressure of 3.0 mbar, i.e. filling nitrogen density of 1.4 × 1017 cm−3, the electron density can be approximately estimated to be 7.2 × 1019 cm−3 at the compressed condition assuming the average charge state of nitrogen to be z = 5 (under-estimated). Due to under-estimation of the compression ratio and the average charge state of nitrogen, the actual electron density may be higher than what is estimated here. Similarly, a filling gas pressure of 8.5 mbar corresponds to an approximate electron density of 2.0 × 1020 cm−3. In this context, it is worth referring to Nevrkla et al. [40] where a temperature of ~90 eV was estimated from spectroscopic simulation with a comparatively larger gain-volume and smaller current driver. Similar case can also be seen in report by Gissis et al. [34] where electron temperature of ~75–85 eV was estimated. Since, comparatively higher and faster electrical driver has been used in our experiments with a smaller gain-volume, it is felt strongly that the temperature reached here should be higher than what was achieved in those reports. However, a detailed spectroscopic simulation will be done in near future to estimate the temperature. This is also well supported by the appearance of fast signal (~2 ns rise time) in the temporal profile at smaller gas pressures. However, gas pressure cannot be reduced beyond a certain limit in order to keep the ion density sufficient enough to produce detectable intensity in Ly-α line. Also, for a fixed current profile, lowering the gas pressure reduces the pinch time of the plasma, as seen in Fig. 3 and if the pinch time becomes substantially lower than the quarter period of the discharge current, the discharge will not be efficient [38]. For an efficient discharge, pinched plasma must be formed at an instant when current nearly reaches its peak value. This means that lower gas pressure can be used with a capillary of higher inner diameter (ID) to compensate for the decrease in pinch time. However, this is also not a good solution because the uniformity of such larger diameter plasma column will be difficult to maintain during the axial compression phase of the plasma. Instead of using higher ID capillary, there is another possibility of making the current faster so that the quarter period of the current (few ns) will go down to a value comparable to the pinch time. Here, of course, one has to work with capillary of smaller ID. In this context, it is worth referring to the recent results by Avaria et al. [41] in which it has been shown experimentally that homogeneous plasma columns with ionization levels typical of mega ampere discharges can be created by rapidly heating gas-filled channels of ~520 μm diameter with nanosecond rise time current pulses of moderate amplitude (~40 kA).

Fig. 6
figure 6

Transmission grating spectrum at 95 kA current and 8.5 mbar gas pressure, in absence of pre-pulse

Full size image
Fig. 7
figure 7

Transmission grating spectrum at 95 kA current and 2.0 mbar gas pressure, in presence of pre-pulse

Full size image

4 Conclusion

Discharge current with a very promising profile having ~95 kA amplitude and 46 ns quarter period was used to excite nitrogen plasma for exploring recombination X-ray lasing. Fast pulse (5–8 ns rise time) was seen with quadrant vacuum diode and was optimized by varying the nitrogen gas pressure in the capillary. The spectroscopic study carried out at these fast and large current excitations showed strong emission of He-α line at 2.8 nm and another spectral line at 2.1 nm, along with various plasma emission lines of different charge states of nitrogen. The Ly-α spectral line of N6+ at 2.4 nm was not visible at higher gas pressures. However, its faint signature could be observed when the gas pressure was reduced significantly, indicating the requirement of more heating of nitrogen plasma for generation of higher charge states, i.e. N6+ and N7+ in sufficient amount. The role of the pre-pulse was also found to be important in getting higher ionization states, indicating the requirement of a stronger pre-pulse. A higher and faster current pulse at smaller gas pressures with smaller diameter capillary may be more advantageous towards demonstration of recombination X-ray lasing at 13.4 nm in nitrogen plasma.