1 Introduction

The development of reflectometry in the soft X-ray (SXR) and extreme ultraviolet (EUV) ranges requires the creation of high-intensity radiation sources. For these purposes, various radiation sources can be used, but laser-plasma ones are considered the most effective [1]. To obtain bright emission radiation in SXR and EUV spectral ranges, it is necessary to form a high-temperature plasma, which includes ions emitting in this spectral range.

There are two options for creating bright radiation sources. These are a source based on a tin drop target [2, 3] and a source based on a xenon gas jet [4, 5].

Xenon is a heavy inert gas, which automatically makes it a convenient target for a laser-plasma source of SXR and EUV radiation. The study of xenon as a target for a laser-plasma radiation source in the SXR and EUV ranges was carried out in a very large number of works, for example, Refs. [6,7,8,9,10]. Usually, in such works, the conversion efficiency of a laser-plasma source with a xenon target at a wavelength of 13.5 nm was investigated. In addition, in a number of works, emission bands have been studied, for example, in the vicinity of 11.3 nm [11, 12] or in the longer wavelength region [13]. At the same time, other emission bands are also of interest for laboratory applications [14], including emission bands in the so-called “water window” 2.3–4.4 nm [15, 16].

The main goal of our investigations was to study the emission spectra of xenon gas-jet targets and the absolute intensities of radiation in the spectral range of 3–20 nm under various conditions of gas outflow and excitation by laser pulses with different parameters.

2 Experimental arrangement and operation

The scheme of the research facility is shown in Fig. 1.

Fig. 1
figure 1

Scheme of the research facility. 1—laser, 2—laser radiation power sensor, 3—dividing plate, 4—prism, 5—optical input, 6—lens, 7—nozzle, 8—quick-acting valve, 9—vacuum shutter, 10—two film free-standing filters, 11—spectrometer-monochromator RSM-500, 12—turbomolecular pump, 13—two film free-standing filters, 14—spectrometer for measuring absolute radiation intensities

Full size image

The operation of the installation is carried out as follows. The gas enters the quick-acting valve 8 and further into the conical supersonic nozzle 7 and then is pumped out by cryocondensation and cryoadsorption pumps. Radiation of the laser 1 falls on the dividing plate 3, from where a small part of the radiation is reflected to the first radiation power detector 2. The main part of the radiation, passing through the prism 4 and the optical input 5, falls on the lens 6. At the focus of the short-focus lens, the laser radiation causes a breakdown and the formation of plasma in a gas jet. Polychromatic soft X-ray radiation of laser spark, passing through the electropneumatic vacuum shutter 9 and two free-standing X-ray filters 10 and enters the RSM-500 spectrometer-monochromator. Next, monochromatic soft X-ray radiation is detected by a pulsed detector. The RSM-500 is pumped out by a turbomolecular pump 12. During the work, we used spherical mirrors and gratings. The radius of curvature of the mirror is 4 m, of the grating is 3 m. The number of grooves is 600 grooves/mm. The spectral resolution of the device, measured at the L-absorption edges of silicon and aluminum, and the K-edge of beryllium free-standing filters, as well as at the zero-order half-width, was 0.04 nm. For the gratings and mirrors used, the operating wavelength range was 1–20 nm.

Free-standing film filters were installed at the entrance to the RSM-500 spectrometer. These filters transmit radiation in the spectral range 4–8 nm (Ti/Be) or 5–20 nm (Zr/ZrSi2) and at the same time effectively absorb the long-wave noise component of the signal. In addition, free-standing filters protect the detector from particles of various nature formed during the operation of the soft X-ray radiation source, which can effectively reduce background noise. The research facility is described in more detail in Ref. [17].

To study the absolute radiation intensities, a calibrated in absolute units spectrometer based on a multilayer X-ray mirror 14 was used. The spectrometer is a φ–2φ goniometer, in which multilayer X-ray mirror is used as a dispersing element. This device works as follows: soft X-ray radiation of a laser spark passes through the input free-standing film filter and falls on the X-ray mirror. In accordance with the Wulf–Bragg condition, radiation with a certain wavelength is reflected from the mirror. The reflected radiation passes through the second free-standing film filter and is recorded by the detector. Spectrum scanning is carried out by rotating (by an angle φ) the X-ray mirror relative to the incident beam, while the detector is rotated relative to the incident beam by a double angle (2φ). The rotation of the mirror and the detector is carried out using a stepper motor, the condition φ–2φ is provided by a gear. Free-standing film filters were installed in the mirror spectrometer. The filters transmit radiation in the spectral range 4–8 nm (Ti/Be) or 5–20 nm (Zr/ZrSi2). The spectrometer and the principles of its operation are described in more detail in Ref. [18].

The absolute radiation intensities for a wavelength of 11.34 nm were additionally measured using a two-mirror monochromator based on two Mo/Si mirrors using free-standing Mo/ZrSi2 film filter. A two-mirror monochromator, its design, operating principles, and the results of measurements obtained with its help are described in Ref. [19]. An SPD-100UV photodiode was used as a detector. The dependences of the radiation intensity on the gas pressure at the nozzle inlet and the distance from the nozzle exit were studied. During the research, a two-mirror monochromator was installed instead of a spectrometer based on a multilayer X-ray mirror.

To excite the gas stream, Nd:YaG laser was used. This laser has the following parameters: wavelength 1064 nm, variable pulse energy up to 0.8 J, pulse duration 5.2 ns, frequency 10 Hz. Laser radiation is focused on a gas target by a lens with a focal length of 45 mm. The calculated diameter of the focal spot is 66 μm. When laser pulses are focused on a gas beam, gas atoms are ionized and a plasma cloud is formed.

For the system of forming a pulsed gas stream, a pulse valve was used. High gas pressure (up to 15 bar) was created at the entrance to this valve. At the valve output, a conical nozzle was fixed. We used a Bosch injector 0 280 158 017 as a gas valve and a conical supersonic nozzle, with a critical section diameter of 500 μm, 5 mm long, and a solution angle of 11°. In the experiments, the duration of the gas pulse was 0.5 ms.

Gas jets formed during the outflow of gas from conical nozzles to a vacuum, in the general case, have a complex spatial structure, determined by the parameters of the gas at the input in the nozzle and the geometric parameters of the nozzle. The tasks of describing the atomic-cluster jets that are formed with the outflow of the condensing gas from supersonic nozzle into vacuum are especially complicated. The gas-dynamic calculation of the structure of such an atomic-cluster target is very complicated and was not carried out in this work.

For a mirror spectrometer, in accordance with Ref. [18], the energy concentrated in the emission line Eline and the number of photons Nline can be determined as follows:

$${E}_{\mathrm{line}}=\frac{4\pi \cdot \alpha V}{\gamma \delta {T}^{2}R}$$
(1)
$${N}_{\mathrm{line}}={E}_{\mathrm{line}}\frac{{\lambda }_{\mathrm{line}}}{hc}$$
(2)

where V is the signal recorded by the detector, in volts. For the researched spectral range, the sensitivity of the amplifier was α = 10–11 C/V; the solid angle of the detector entrance window as seen from the point EUV radiation source γ = 5.45∙10–5 sr; sensitivity of the photodiode \(\delta\) = 0.25 C/J. T is the transmission coefficient of the Ti/Be film filter on the researched wavelength, and R is the X-ray mirror reflection coefficient at the researched wavelength.

In a two-mirror monochromator, two Mo/Si mirrors are used and, accordingly, the value of R2(λ) has the form of a very sharp function with a peak in the region of 11.34 nm, which makes it possible to determine the energy concentrated in the emission line E11.34 and the number of photons N11.34 as follows:

$${E}_{11.34}=\frac{4\pi \cdot \alpha V}{\gamma \delta {T}^{2}{R}^{2}}$$
(3)
$${N}_{11.34}={E}_{11.34}\cdot \frac{{\lambda }_{11.34}}{hc}$$
(4)

where V is the signal recorded by the detector, in volts. For the researched spectral range, the sensitivity of the amplifier was α = 10–11 C/V; the solid angle of the detector entrance window as seen from the point EUV radiation source γ = 2.32∙10–4 sr; sensitivity of the photodiode \(\delta\) = 0.25 C/J. T is the transmission coefficient of the film filter on the researched wavelength T = 0.5, and R is the X-ray mirror reflection coefficient at the researched wavelength R = 0.21.

3 Experimental results

The xenon emission spectra measured at a gas pressure at the nozzle inlet of 5 and 8 bar and a temperature of 300 K are shown in Fig. 2. A conical supersonic nozzle with critical section diameter 500 µm was used. The range of 3–8 nm was investigated, the laser pulse energy was 0.8 J, and the radiation intensity is given in relative units. Two free-standing Ti/Be filters were used as filters.

Fig. 2
figure 2

Emission spectrum of a xenon gas-jet target at different gas pressures at the nozzle inlet

Full size image

In the spectrum, we can see a number of broad bands in the range of 3–8 nm and a dip in the region of ~ 4.35 nm. This is a property of the xenon plasma itself and is not related to the absorption of radiation by carbon films.

The emission spectra of xenon measured using a capillary d = 500 µm and a cone nozzle with critical section diameter 500 µm at a gas pressure at the nozzle inlet of 3 bar and a temperature of 300 K are shown in Fig. 3. The range of 6–20 nm was investigated. The intensities are given in relative units. Two free-standing Zr/ZrSi2 filters were used as filters.

Fig. 3
figure 3

Xenon emission spectrum when using a capillary diameter 500 µm and a cone nozzle with critical section diameter 500 µm

Full size image

In the spectral range of 8–18 nm, we see a number of high-intensity bands. The most intense is the UTA band (10–12 nm), studied in many works, for example, [20, 21]. From the figure, you can see that the use of a supersonic nozzle leads to a significant change in the intensity of the UTA band, associated with a change in the absorption of radiation in the gas jet and the residual gas (xenon) contained in the chamber [22]. In general, the recorded spectra of xenon in the range of 10–15 nm are extremely dependent on the absorption of radiation in the gas jet and in the residual gas (xenon), which leads to a strong dependence of the observed spectra on the structure of the target jet. To find the optimal parameters for the operation of this source of SXR and EUV radiation, it is necessary to develop a gas target individually for the research facility. Such works are predominantly experimental, since the calculation of supersonic flows with gas condensation is a very difficult task.

We note separately that the type and intensity of xenon emission spectra depend on the relative position of the nozzle exit and the laser spark. The maximum radiation intensity is achieved when the laser spark is located not along the axis of the nozzle, but with a certain displacement depending on the structure of the outflowing jet [11, 23].

Based on the results of the studies, we compiled a table of the observed xenon lines at a gas pressure at the nozzle inlet of 5 bar and identified them. Identification was carried out in accordance with Refs. [24, 25] and is given in Table 1. Xenon is a well-studied gas in the region of 10–18 nm, but in the shorter wavelength range, the identification of emission lines is very difficult.

Table 1 Xenon emission lines
Full size table

We have determined that the registered emission spectra are formed by transitions on Xe–VIII, Xe–IX, Xe–X, and Xe–XI ions, merging into continuous bands. Separately, it is worth noting the UTA band formed by ions from Xe–X to Xe–XIV [20].

The relative line intensities for the spectra measured using various target formation systems are given in Table 2. The line intensities are normalized to the intensity of the Xe–IX (8.84 nm) 4d10–4d97p line. This line is located in the shorter wavelength range relative to the broad xenon absorption band. Therefore, relative to it, we can qualitatively trace the transformations of the spectrum associated with the absorption of radiation in the residual xenon.

Table 2 Relative intensities of xenon lines for various target formation systems
Full size table

From the data in Table 2, we can see that the use of supersonic nozzles for the formation of xenon gas-jet targets makes it possible to obtain significantly lower absorption in the residual gas. Increasing the gas pressure at the nozzle inlet also leads to an improvement in the (source intensity)/(self-absorption) ratio, especially for the 11–17 nm range. For further research, we decided to use conical supersonic nozzles at increased xenon pressure.

Figure 4 shows the emission spectra of xenon, measured at a gas pressure at the nozzle inlet of 2–10 bar and a temperature of 300 K. The diameter of critical section of the nozzle is dcr = 500 μm. The range of 8–18 nm was investigated, radiation intensities are given in relative units.

Fig. 4
figure 4

The emission spectra of xenon at different gas pressure at the nozzle inlet

Full size image

In spectral range of 8–18 nm, we observe a number of intense emission bands. The brightest is the UTA band (10–12 nm), but there are other bright band, in particular 13.5, 14 and 15 nm. You can see that with an increase in pressure, the intensity of the UTA band increases rapidly at first, then the intensity of the bands in the range of 12–16 nm gradually increases. With a pressure of more than 8 bar, the intensities of the bands practically does not increase.

Then, we studied the absolute radiation intensities of the xenon gas-jet target as a function of the gas pressure at the nozzle inlet and the energy of laser pulse.

The emission intensity in absolute units was determined in accordance with the procedure described in Sect. 2. Absolute intensities for emission bands 11.34 ± 0.25 nm, 13.8 ± 0.38 nm and 14.8 ± 0.4 nm per laser pulse into a hemisphere are given in Table 3. Laser pulse energy was 0.8 J.

Table 3 Radiation intensity of the xenon target as a function of the gas pressure at the nozzle inlet
Full size table

The measurements were carried out using a Cr/Sc mirror and two Ti/Be film filters at a wavelength of 4.4 nm, for a wavelength of 10.8 nm with a Mo/B4C mirror and two Zr/ZrSi2 film filters, for a wavelength of 11.34, 13.8, 14.8 nm with Mo/Be mirror and two Zr/ZrSi2 film filters. The data from Table 3 are also presented graphically in Fig. 5.

Fig. 5
figure 5

Dependence of the absolute radiation intensities of the xenon target on the gas pressure at the nozzle inlet

Full size image

You can see that as the pressure increases, the radiation intensity reaches saturation at a pressure of about 8 bar. Thus, for the equipment we used, we chose a nominal pressure of 8 bar.

Then, we studied the radiation intensity of the xenon-based target as a function of the laser pulse energy. Figure 6 shows the xenon emission spectra measured at a laser pulse energy of 0.2–0.8 J, the gas pressure at the nozzle inlet was 8 bar, and the temperature was 300 K. The range of 8–18 nm was studied, and the radiation intensities are given in relative units.

Fig. 6
figure 6

Emission spectra of xenon at different laser pulse energies

Full size image

You can see that initially the intensity of the bands corresponding to ions with low ionization degrees (Xe–VIII) increases, then the intensity of the UTA band increases and further the intensities of the bands corresponding to ions with high ionization degrees (Xe–XI) increase.

Table 4 gives the absolute intensities for the emission bands 11.34 ± 0.25, 13.8 ± 0.38 and 14.8 ± 0.4 nm per laser pulse per hemisphere at various laser pulse energies.

Table 4 Absolute radiation intensities of the xenon target at a gas pressure of 8 bar at the nozzle inlet and various laser pulse energies
Full size table

The data from Table 4 are also shown graphically in Fig. 7.

Fig. 7
figure 7

Dependences of the absolute radiation intensities of the xenon target on laser pulse energy

Full size image

You can see that the emission intensity of the 11.34 ± 0.25 nm band is significantly higher than for other emission bands. The radiation intensity depends approximately linearly on laser pulse energy. Thus, the use of lasers with a reduced pulse energy makes it possible to efficiently excite a xenon gas-jet target.

In addition, we studied the dependence of the emission intensity for the band 11.16 ± 0.13 nm on the distance from the nozzle exit at various gas pressures at the nozzle inlet. The studies were carried out using a two-mirror monochromator based on multilayer Mo/Si mirrors using two Mo/ZrSi2 film filters. Absolute radiation intensities per laser pulse per hemisphere are given in Table 5. Laser pulse energy is 0.8 J.

Table 5 Absolute radiation intensities of the xenon target as a function of the distance between laser spark and the edge of nozzle at various pressures
Full size table

The data from Table 5 are also shown graphically in Fig. 8.

Fig. 8
figure 8

Dependences of the absolute intensities of radiation of a xenon target into a hemisphere on the distance between the laser spark and the nozzle exit at various gas pressures

Full size image

From Fig. 8, you can see that the radiation intensity decreases rather slowly as the laser spark moves away from the nozzle exit, which makes it possible to work at large distances from the nozzle exit. Thus, it is possible to significantly reduce the degradation of the nozzle during the operation of the laser-plasma source.

4 Conclusion

As a result of the studies of xenon emission spectra, we can make the following conclusions:

  1. 1.

    For the studied types of gas targets, the form of the recorded xenon spectrum changes qualitatively. These changes are explained by the high level of self-absorption of emission radiation both in the studied jet and in the residual gas in the vacuum chamber. Thus, the recorded emission spectra can vary significantly depending on the gas pressure, the gas outflow regime, the vacuum pumping power, and the geometrical parameters of the research facility.

  2. 2.

    The use of supersonic xenon gas jets makes it possible to increase the radiation intensity of the laser-plasma source. In this case, the gas jet will be collected in a smaller solid angle and will have a higher density. The concentration of the residual gas in the vacuum chamber decreases.

  3. 3.

    The intensity of the emission radiation in the range of 8–18 nm when using xenon is very high. When measuring on a two-mirror monochromator at a gas pressure at the nozzle inlet of 10 bar, a laser pulse energy of 0.8 J, and x = 0, the conversion efficiency of laser radiation energy into EUV radiation in the 11.16 ± 0.13 nm band was obtained at a level of CE = 1%. In general, xenon is a very promising target gas. But due to the large influence of self-absorption, it is necessary, if possible, to reduce the gas flow and use more efficient pumping systems.

  4. 4.

    The efficiency of conversion of laser radiation with a long wavelength of 1064 nm to the EUV obtained by us can be compared with the results obtained when plasma was generated using shorter wavelength laser radiation. In Ref. [26], a liquid Xe jet was excited by a Nd:YaG laser at the second harmonic of 532 nm at a wavelength of 11.3 nm in 1.4% BW; CE = 0.4% was obtained. In Ref. [27], the Xe gas jet was excited by a KrF laser with a wavelength of 248 nm. At a wavelength of 12.8 nm in the band, 1% received CE = 0.2%. Thus, you can see that the conversion efficiency obtained by the authors in this work is higher than when plasma is generated by lasers with a shorter wavelength.

  5. 5.

    In the spectral range of 3–8 nm, broad emission bands are observed, but the absolute xenon emission intensities are low. For this spectral range, xenon is an unpromising target.

As a result of the work carried out, high intensities of emission radiation in the SXR and EUV spectral ranges were obtained using xenon gas-jet targets. It is possible to develop radiation sources, based on such systems, for industrial use in lithography.