Laser produced plasma source (LPP)

The LPP soft X-ray source is pumped by a homemade CPA thin disc laser (TDL) system. The TDL system consists of a front-end (YGW oscillator and regenerative amplifier), a stretcher and a Yb:YAG thin disc regenerative amplifier (cp. Fig. 1) The output of the regenerative amplifier is compressed using a grating compressor to a pulse duration of 1.6 ps. The TDL operates at 1030 nm with 100 Hz repetition rate and delivers pulses with an energy of 120 mJ after compression. In our experimental setup the laser is focused with a f5 optic at an angle of ~45° on a metallic rotating target cylinder. In the focal region with a FWHM of 17 μm we reach 60% energy content corresponding to an intensity of about 2x10¹⁶ W/cm² on the target. The plasma created by the laser pulse emits a wide soft X-ray spectrum. We measured the soft X-ray source size with a pinhole camera setup to be at least 2x the laser focal size (FWHM) from which the incoherent soft X-ray radiation (SXR) is uniformly emitted. For debris protection a thin glass plate after the entrance window of the vacuum chamber and an optional 900 nm Mylar foil after the target on the entrance of the soft X-ray beamline were used. Fig. 1 shows the scheme of the laser and beam guiding system in our setup.

Figure 1.

Schematic of the LPP source setup including the CPA thin disk pump laser

High Harmonics Generation (HHG) source

The HHG soft X-ray source is driven by a state-of-the-art high power OPCPA system [2]. A schematic of the setup is shown in Fig. 2. The high average power OPCPA system is based on a single Yb:YAG thin disk laser (Dira 500-10, Trumpf Scientific Lasers) delivering 500 W average power at a pulse repetition rate of 10 kHz. This output is used both as pump and signal generation source for the OPCPA system. For this purpose, it is split into two arms. The low power arm contains 0.5 mJ pulses, compressed to below one picosecond, which are delivered to a front-end manufactured by Fastlite. The high power arm contains the remaining power, compressed to a pulse duration of about 2 ps. In the frontend, the seed is obtained by super continuum generation and subsequent difference frequency generation. Two PPLN-based OPA stages are used as pre-amplifiers, and an acousto-optic programmable dispersive filter serves as pulse-shaper.

The output pulses of the front-end have an energy of about 2.7 μJ and extend from 1800 nm to 2600nm, with a central wavelength of 2.1 μm. In the subsequent power amplification, the pulse energy is boosted to more than 3 mJ using two home-built OPCPA stages. The first stage is pumped with 130 W average power and amplifies the 2.1 μm pulses to about 1 mJ using BiBO as nonlinear crystal. In the second stage, a YCOB crystal pumped with 320 W is used to increase the pulse energy to its final value. A part of this power is separated to be used as pump beam in pump-probe experiments, so that 2.8 mJ are available for HHG. The pulses are compressed using a 41 mm thick suprasil glass block, placed under an angle in the beam such that it compensates the dispersion accumulated in the setup. The compressed pulse duration typically is about 26 fs, with a spectral width of more than 400 nm FWHM.

Figure 2.

Schematic of the OPCPA and HHG setup.

The compressor serves at the same time as entrance window in the vacuum setup used for HHG. The beam is focused by a 500 mm lens into a differentially pumped 5 mm long gas cell filled with up to 1.5 bar of helium. Behind the cell, most of the transmitted IR radiation is reflected by a mirror, where the SXR radiation can pass through a hole in its center. The remaining IR light is blocked by one or more thin foil filters, so that only the X-rays pass through a slit and a transmission grating onto an X-ray CCD camera.

Photon Flux Evaluation

Both soft X-ray sources were characterized using a transmission grating spectrometer (TGS) setup schematically shown in Fig. 3). It consists of a 10.000 l/mm free standing Si₃N₄ transmission grating [6] and an IDus 420 CCD detector (ANDOR, 1024x255 pixels). Thin metal foils (Al, Zr) were used to shield the IR light. The foils also provide absorption edges for the spectral calibration. The detector was mounted on a rail that allowing to change the detection angle in the range of ±10 degree. Our setup provided a detection angle of 2.1x10^-8 sr with a slit 3 mm in height and 50 μm in width, that was placed just before the transmission grating. The setup was calibrated at the German national metrology institute, Physikalisch-Technische Bundesanstalt (PTB), and can therefore be used to account for the absolute number of photons generated in the LPP and HHG process.

Figure 3.

Calibrated transmission grating spectrometer (TGS) used for the evaluation of the photon flux of the LPP. For the evaluation of the photon flux of the HHG a similar setup (not shown) has been used.

Reflection zone plate spectrometer

For our NEXAFS investigations with the LPP shown below we used a spectrometer setup with reflection zone plates (RZP) on planar substrates. It is described in more detail in [5].

In order to further improve spectrometer efficiency, spectral resolution and energy range in NEXAFS investigations we rely on a reflection zone plate array (RZPA) on a spherical substrate. The main references concerning a possible optical layout solution is the RZPA spectrometer as published in [7, 8]. Recent advancements with RZPAs on a spherical substrate are reported in [9]. As an option, the RZPA provides up to four different RZPs, each with the same resolving power, together covering the energy range (150 - 700) eV with HHG source and (210 - 1400) eV with LPP source. The spectrometers provide a single-shot online monitoring of the initial spectra of the source, energy resolving power E/AE above 1000 and peak efficiency up to 25%. As a detector typical X-ray CCD or sCMOS camera with pixel size of 13.5 μm × 13.5 μm can be used. We have projected two related spectrometer designs for the LPP and HHG source respectively. As an example Fig. 4 shows the optical layout of the proposed LPP spectrometer.

Figure 4.

Optical layout of the LPP spectrometry system. See Table 1 for the notation and data.

The X-ray beam from the LPP source is transmitted through a sample, which is located in a distance of 0.5 m from the target. The parameters of RZPAs in reference and signal channels are identical. The position of the CCD detector is assumed to be fixed. Therefore, in order to cover the full energy region of interest four RZPs are designed on the same substrate. Parameters of these RZP structures are shown in table 2. Each RZP has an aperture of 6 mm sagittal and 80 mm meridional fabricated on a spherical substrate with a radius of 45 m.

Table 1.

Optical layout parameters.

Table 2.

Parameters of the optical elements within RZPA (signal and reference channel)

The results of the calculation of expected efficiency and energy resolving power are shown in figure 5a and 5b accordingly.

Figure 5a.

Calculated RZP efficiencies, assuming Au coating with a CO contamination layer (2 nm) on the top of Au.

Figure 5b.

Energy resolving power of RZPs 1 – 4, covering the energy range (210 – 1400) eV.

Photon flux and spectra of the LPP source

Using the calibrated TGS setup (cf. Fig. 3) soft X-ray spectra were measured with different target materials copper and tungsten. For the evaluation we averaged datasets of at least 10x10 s at the laser’s repetition rate of 100Hz. Fig. 6 shows as an example the measured spectra emitted by a tungsten target and filtered by Zr (298nm thickness) or Al (200nm) foils. Characteristic absorption edges by either filter (Al, Zr), filter contaminations (O, C) and the supporting structure of the transmission grating (Si₃N₄) can be identified in the tungsten spectra. The photon flux of the LPP source was evaluated from the averaged raw data (counts/s) with the efficiency of the transmission grating (theoretical efficiency function normalized to a fixed calibrated photon energy), filter transmission characteristics (calculations using Henke database [10]) and the calibrated CCD detector. Fig. 7a presents the resulting flux in photons/s*sr in 0.1% bandwidth for the tungsten target and Fig. 7b for selected Cu-emission lines. Both figures show, that the LPP source emits in a wide spectral range (50 – 1500 eV) at least 10¹¹ photons/s *sr @0.1% bandwidth.

Figure 6.

Measured emission spectra of the LPP source (left) filtered by Zr and Al foils (right) Target: tungsten.

Figure 7.

Photon flux of the LPP source corrected for filter transmission, grating efficiency and detector response for the tungsten (a) and the copper target (b).

Photon flux of HHG

Two typical spectra recorded with our HHG source can be seen in Fig. 8. The green curve was taken with only an aluminum foil filter, whereas for the purple curve, an additional titanium foil was inserted. As can be seen from Fig. 8 the Ti L-edge is clearly visible. Using the calibrated spectrometer setup described above, the foil filter transmission curves from [10] and an estimation of the ratio of photons passing through the spectrometer entrance slit, we calculated a photon flux of up to 10⁶ photons/eV/s at a photon energy of 450 eV. The obtained flux is comparable to already reported values obtained with laser drivers of lower repetition rate [11-13] but the achieved single pulse conversion from IR to SXR energy is lower and needs further investigation. One has to admit that, in contrast to our measurements, the majority of reported values is not based on calibrated measurement devices.

Figure 8.

Typical HHG spectrum obtained with the transmission grating spectrometer. Gas medium: He at 1.8 bar. Laser parameters: 9x10¹⁴ W/cm², 25 fs, 10 kHz. Filters: Al – 200 nm, Ti - 200 nm.

X-ray absorption (XAS) investigations

We used the TGS spectrometer (cp. Fig. 3) and a spectrometer setup based on plane RZP substrates [5] to demonstrate the capability of both sources for NEXAFS investigations on thin metal films. Fig. 9 shows NEXAFS a spectrum of a Ti foil (200 nm) registered with the TGS and the HHG source in comparison with the spectrum of the same sample registered using the LPP source and a spectrometer with plane RZP substrates. As can be seen from Fig. 9 both the resolution (cp. edge jumps of both spectra) and the signal to noise ratio of the spectrum could be significantly increased in the case of the RZP spectrometer. Nevertheless in both spectra characteristics features of the Ti L-edge absorption are visible. The broad spectrum of the LPP source together with the highly efficient RZP spectrometer provides the possibility to investigate also extended X-ray absorption fine structures (EXAFS) at L-edges of transition metals. As an example Fig. 10 shows the XAS spectrum of a Ni foil covering a large photon energy range. EXAFS oscillations are clearly visible.

Figure 9.

NEXAFS spectra of a Titanium foil (200 nm) measured with a HHG source and a TGS spectrometer in comparison with a spectrum recorded using the LPP source and a spectrometer based on plane RZP substrates.

Figure 10.

L-edge XAS spectrum of a Ni foil (200 nm) showing both near edge structures as well as oscillations in the EXAFS region. The spectrum was recorded with a RZP spectrometer based on plane substrates.