Open Access
Add to BibSonomy
Abstract
In this study, we demonstrate the significant increase in the hard X-ray yield (more than 1011 photons/s in 4π solid angle in 6 - 40 keV range) that can be achieved in an ambient air environment when solid targets are irradiated by sequences of high average power (90 W) bursts of femtosecond laser pulses, generated in GHz burst laser amplifier operated at high repetition rate (100 kHz). The combination of the prepulse and ∼ 10 times greater driving pulse not only enhances X-ray generation efficiency (∼ 10−6) by more than two orders of magnitude compared to the single pulse regime but also protects a target allowing continuous operation for 3 hours with only 30% predictable and gradual drop of X-ray yield. In addition, we show that X-ray yield enhancement becomes around 6 times more pronounced at higher repetition rates (100 kHz compared to < 5 kHz). The simplicity and relative cost-effectiveness of the presented X-ray source makes it an attractive solution for future applications in ultrafast X-ray imaging and spectroscopy.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
X-rays are a well-established tool of scientific research in areas like physics, chemistry, biology, and medicine. In recent decades, X-ray technologies have matured to the point where steady-state X-ray diffraction, imaging and spectroscopy equipment have become a part of the standard toolkit of research labs. In addition to research applications, soft X-ray (SXR) and extreme ultraviolet (EUV) spectral range is essential for their applications in the semiconductor industry, where the dimensions of lithographic features are approaching the X-ray wavelengths.
While these steady-state X-ray methods are well-established, femtosecond and picosecond time-resolved X-ray crystallography, imaging and spectroscopy remain in the realm of large-scale synchrotron facilities because they are the only femtosecond sources capable of producing ultrashort X-ray pulses of sufficient brightness. Since access to such facilities is limited, there is an apparent interest in the development of compact, high average flux, reliable, stable, cost-effective X-ray and EUV sources as an alternative.
Laser-produced plasma (LPP) X-rays are one of the alternatives, which covers the EUV, SXR, and hard X-ray (HXR) range [1]. Ultrafast LPP X-ray radiation with photon energies in range of 2–20 keV finds applications in fundamental structural dynamic studies of materials [2], in particular X-ray absorption spectroscopy (XAS) [3–5] and X-ray diffraction (XRD) [6–9]. Laser-driven XAS in the range of 2 - 6 keV is employed to study warm dense matter [10].
In addition to the generating short pulses, LLP sources are also capable of producing a high degree of spatial coherence as a result of their small spot size. The size of the radiating plasma on the laser intensity typically is just a few times larger than laser focal spot [11], which enables applications in X-ray micro computer tomography (micro-CT) systems [12] or phase-contrast imaging applications [11,13,14]. For these imaging applications, low photon absorption is desirable; typically, they require harder X-rays in 10 to 100 keV range.
Usually, ultrashort pulse LPP X-ray generation is achieved using ps-fs laser pulses in near-IR. The femtosecond X-ray flash is generated when an intense femtosecond laser pulse interacts with a target creating high-temperature electron plasma [15,16]. The emission of such plasma consists of Bremsstrahlung and characteristic X-ray lines of the target material with Kα and Kβ lines most pronounced. In transmission geometry, when driving pulse is hitting the target surface opposite to that from which the X-ray photons are collected, the X-ray pulse duration is comparable with that of the laser pulse producing the plasma [1,17]. In reflection geometry, when X-ray photons are collected from the same side of the target where the excitation beam impinges, the X-ray pulse duration is expected to be several picoseconds even when driving with fs pulses [18]. In particular, the rise time of the X-ray emission is mainly determined by the characteristic time needed to heat the plasma, while the time scale of plasma cooling sets the fall time of the emission. The primary mechanisms of plasma cooling are plasma expansion and radiation emission.
The energy of produced X-ray photons is directly related to the electron temperature (Te), which is a function of laser intensity (${I_L}$) and wavelength (${\lambda _L}$) and could be written as $\; {T_e} \propto {I_L}\mathrm{\lambda }_L^2$ [15]. Moreover, X-ray yield (${Y_{X - ray}}$) is also intensity-dependent (${Y_{X - ray}} \propto {({{I_L}} )^a}$), where $\alpha $ value depends on the material and could be even higher than 2.5 [19]. It suggests that a straightforward method to obtain more photons from LPP is to use significantly more powerful pulses, which might exceed even 1021 W/cm2 intensity on a target [20]. Unfortunately, laser field intensities higher than ∼ 1014 W/cm2 are clamped by ionization of ambient air [21,22]. The intensity clamping threshold can be increased up to 3 times by placing the target in the atmosphere with higher ionization potential, such as helium. However, to completely avoid clamping and exploit all the intensity available from modern femtosecond lasers, the experiments must be performed in vacuo. Besides making the entire setup significantly more complex and harder to deploy, using vacuum creates additional problems: the debris and particles ablated from the target by laser pulse are readily deposited on the beam steering optics, quickly yielding them unusable without special precautions [1,23–25].
As an alternative to the brute-force increase of pulse energy, X-ray photon flux can be increased using lasers featuring moderate pulse energies but operating at higher pulse repetition rates [26]. It allows working in ambient air atmosphere still delivering high flux values (∼ 109 photons/s into 4π solid angle) with a relatively high process conversion efficiency (∼ 10−7) [27]. Moreover, ytterbium (Yb) based lasers with high repetition rates are more readily scalable in average power than titanium sapphire (Ti:sapphire) based systems [28], which have traditionally been used for LPP X-ray generation.
Specially prepared targets with micro or nanostructured surfaces have also been suggested as means of increasing X-ray yield [29–32]. The structure on the front surface may significantly boost the absorption and increase the electron temperature and density, which leads to higher ion acceleration efficiency and maximum ion energy [33]. In some cases, such structures can be made by the same femtosecond laser in situ by multiple prepulses irradiated prior to a subsequent much more intense mail laser pulse [34]. An important variant of such in situ target preparation is the application of a (relatively) weak prepulse immediately before the main laser pulse to modify the interaction between the targets with the intense laser pulse [6,35–37]. The prepulse does not provide X-ray emission; however, it creates some gaseous plasma. By controlling the delay and amplitude of the prepulse, one can control the properties of the plasma found by the main pulse, thus controlling the main pulse coupling efficiency. It has been shown that introducing such prepulse increases the X-ray yield up to 100 times compared with the single pulse operation [38–40]. Moreover, electron beams with energy peaked in the range of 280–390 keV might be ejected using the prepulse and driving pulse approach [41]. These can significantly impact both the applications or the secondary radiation and have implications for the safety of the experimenter.
This work presents experimental results demonstrating a significant increase (more than two orders) in the hard (6–40 keV) X-ray yield that can be achieved in an ambient air environment if various metal targets are irradiated by a pair of high average power (90 W) two femtosecond laser pulses (with 440 ps delay), generated in GHz burst laser amplifier at high repetition rate (100 kHz). X-ray flux dependence on focusing conditions, prepulse amplitude, and prepulse delay time were evaluated at various repetition rates. We also present long-term X-ray emission measurements in single and double pulse regimes. Finally, we show proof-of-principle applications of such X-ray source in imaging and spectroscopy.
2. Experimental setup
Figure 1(a) shows the scheme of the experimental X-ray generation and detection setup. Laser radiation is delivered by solid-state femtosecond laser (Carbide, Light Conversion). Its average power output is 90 W, the minimum pulse duration is 240 fs (FWHM), a repetition rate can be adjusted from single-shot to 2 MHz (maximum average power and maximum pulse energy are achieved simultaneously at 100 kHz), and the central wavelength is 1030 nm. The laser beam quality parameter M2 was less than 1.2, and the focused laser beam spot size diameter was 25 µm (FWHM), resulting in ∼ 450 TW/cm2 driving intensity onto the target surface at 0.9 mJ pulse energy. Linearly polarized laser light was focused (f = 150 mm) and reflected by a dielectric mirror onto the rotating disk target surface at 0° incidence angle, where LLP X-ray generation took place. The lens was mounted on a motorized translation stage system (Standa 8MT160-300) to be able to explore the influence of focusing conditions.
Fig. 1. Experimental setup and typical X-ray spectrum. (a) Schematic layout of LPP X-ray generation and detection setup. (b) Schematic layout of prepulse generation with a variable delay time. BS: beam splitter; M: dielectric mirror; DL: delay line; λ/2: half-wave plate; BD: beam dump; PL: polarizer. (c) Typical spectrum of an X-ray radiation plasma source obtained d = 220 cm from the source for 60 seconds via Amptek X-123SDD spectrometer, with laser pulses of 100 kHz repetition rate and 0.9 mJ pulse packet energy (EPP = 0.08 mJ, EDR = 0.82 mJ and FPP = 11.8 J/cm2, FDR = 118 J/cm2 fluence on the target surface) using Fe, Cu, and Al targets and Al 22 µm foil before the spectrometer.
Prepulses were generated using two methods. The first method employed the GHz/MHz burst mode implemented in the laser [42]. In burst mode, the laser output consists of pulse packets instead of single pulses. Each packet consists of a certain number of equidistant pulses with the same polarization. MHz-Burst contains from 1 to 10 pulses with a fixed nanosecond period (13.15 ns), GHz-Burst contains 1 to 10 pulses with a fixed picosecond period (440 ps). If both bursts are used, the equally separated nanosecond pulse packets contain picosecond sub-packets of pulses. Moreover, the distribution of pulse amplitudes within bursts can also be adjusted via the laser settings. To evaluate pulse amplitudes and spacings within the burst, we used a time-correlated single-photon timing module (Becker&Hickl, SPC-140) with an actively quenched single-photon avalanche diode (SPAD) (ID Quantique id100-20) as a detector, providing the time resolution of 50 ps.
Since the delay between the pulses within the laser-produced bursts is fixed, we have used an additional pulse delivery scheme to generate the prepulse with variable delay. The single laser pulse was split into two; one of the pulses was delayed in the optical delay line (hollow corner cube on a motorized translation stage) and recombined with the main pulse before arriving at the target (see Fig. 1(b)). In both branches, attenuators (λ/2 plate and a polarizer) were placed to control the amplitude ratio of the two pulses. Besides higher complexity, this setup leads to ∼ 40% energy losses and 0.55 mJ total energy of the pulse packet (EDR = 0.5 mJ and EPP = 0.05 mJ). It is important to note that polarization and focusing conditions to the prepulses and driving pulse were the same using both methods.
The target was based on a rotating disk geometry, which was continuously translated along the radial direction in the focal plane, and the laser pulse angle of incidence was kept at zero. The disk rotation speed was ∼ 800 RPM, corresponding to the linear velocity of 4.2 m/s at the inner diameter (5 cm) and the 8.4 m/s at the outer diameter of the target (10 cm). In contrast, the lowest calculated speed required for each consecutive laser pulse separation was ∼ 4.5 m/s at 100 kHz. Employing such target delivery configuration, LPP X-rays were generated and measured using Al, Fe, and Cu metal disk targets.
Direct measurements of the X-ray emission spectra from plasma with an energy resolution of 140 eV and 50 eV channel width were done using an Amptek X-123SDD spectrometer where the detector is protected by 25 µm Be foil. To attenuate the X-ray signal, the different thickness Al foil filters were used. These filters allowed us to easily attenuate X-ray flux under 6 keV, minimizing the total spectrometer input count rate below 10% of applied laser pulse repetition rate and avoiding possible detector saturation and photon pile-up phenomena. The measurements were performed in an ambient air environment, with no additional shielding gas (such as helium) to the target, using only suction (shop vacuum) for debris minimization near the target. The distance between the spectrometer and the target was adjusted depending on the X-ray flux being measured. At the highest flux values, the distance d was 220 cm, with θ ≈ 10° angle to the target plane (see Fig. 1(a)) and 22 µm Al foil before the detector. An example of measured spectra is presented in Fig. 1(c). To calculate the number of generated photons, the transmission of the filters, position of the detector, and detector efficiency were taken into account. The angular distribution of X-rays from the laser-driven plasma sources was assumed to be almost isotropic [43]. X-ray fluxes in desired photon energy range (Eph1 – Eph2) were calculated using Eq. (1):
3. Results and discussion
3.1 LPP X-ray source optimization
We first discuss the experiments directed to optimize the yield of LPP X-ray source. The main three parameters investigated to maximize X-ray yield were (1) focusing conditions, (2) amplitude ratio between prepulse and driving pulse at fixed pulse packet energy, and (3) delay time of the prepulse. The first two dependencies were measured using laser burst mode at 100 kHz and various pulse packet energies. A number of different MHz and GHz laser burst configurations were investigated; however, we found that maximum X-ray yield was generated using only one ∼ 440 ps prepulse (GHz burst mode). Obviously, the number of different material and pulse configurations is virtually unlimited, and we only checked the most obvious ones (ps and ns pulse triplets with different pulse energy ratios within the bursts). We note that some interesting results have been reported showing a high increment of the X-ray emission during material processing at various MHz and GHz laser burst configurations at multiple repetition rates and scanning conditions [44,45].
Figure 2 presents the registered X-ray emission in the range of 6–40 keV as a function of relative distance change between the focusing lens and the target. These dependencies were measured at various pulse packet energies when 440 ps delayed prepulse, and the driving pulse amplitude ratio was 1/10, respectively.
Fig. 2. X-ray emission in 6–40 keV range dependence on the lens distance change at different pulse packet energies, when prepulse and driving pulse amplitude ratio was 1/10 respectively, and repetition rate was 100 kHz. The dotted red line represents how the optimal lens distance depends on the energy of the pulse packet.
It can be seen that the optimal distance from the lens to the target mainly depends on pulse energy and decreases with increasing intensity (positive x-axis values represent increasing distance). The dashed red line in Fig. 2 illustrates how optimal lens distance changes from the energy of the pulse packet. The distance change is ∼ 1.2 mm in the range from 0.22 mJ to 0.9 mJ of pulse packet energy. This dependence was also observed by other authors [14] and can be explained by the phenomena of the self-focusing in the air [46] and air clamping effects [21] when the distance of focused beam intensity maximum decreases with the increasing power of a pulse. The critical self-focusing power for 1030 nm wavelength pulse is around 3 GW in air, and the power of 240 fs, 0.9 mJ pulse is ∼ 3.5 GW, which exceeds this critical power limit. On the other hand, the air clamping intensity is around 100 TW/cm2, which is reached with 0.2 mJ pulse energy at our experimental conditions. Therefore, these two effects may influence the optimal target distance from the lens for maximal X-ray generation. We also noted that vacuum suction near the plasma spot not only increases registered X-ray signal (presumably, by limiting reabsorption and scattering losses due to debris), but also reduces the optimal target distance from the lens decreases by ∼ 1.5 mm when suction is removed. This change is probably caused by the reduction in ionization potential when particles are present in the air around the target. In further experiments, the position of the suction nozzle with respect to the excitation spot was fixed and the optimal focusing conditions were found before each experiment.
Another parameter to optimize for the most optimal X-ray generation is the ratio of the main pulse and prepulse amplitudes. The experimentally measured dependence of the generated X-ray radiation on the ratio of the main pulse to the prepulse energy is shown in Fig. 3(a). It is important to note that the amplitude ratios were evaluated using SPAD module measurements (see Fig. 3(b)), simply dividing the registered signal amplitude of the prepulse and the driving pulse. SPAD module measurements also showed that the time delay between the pulses in a burst was fixed at around 440 ps (as dictated by the construction of the laser).
Fig. 3. (a) X-ray emission in 6–40 keV range dependence on the prepulse to the driving pulse energy ratio, at 0.9 mJ, 0.6 mJ, and 0.18 mJ total energy of the pulse packet and 100 kHz repetition rate. (b) The SPAD measurement of energy ratio between prepulse and driving pulse.
As can be seen in Fig. 3(a) an optimal pulse energy ratio results in the increase of X-ray yield by more than two orders of magnitude when compared to a single pulse operation. If the prepulse energy is too low, no prepulse effect, such as plasma formation, could be initiated. If the prepulse energy is too high, the generation efficiency drops due to the lower energy available for the main pulse. The window of efficient generation is quite broad: 3% to 30% of the driving pulse energy should be delivered to the prepulse. The optimal value of prepulse to driving pulse energy is around 10% at all investigated pulse energies. Since this number may depend strongly on the experimental conditions (ambient pressure, target material, focusing, etc.), it is important to be able to adjust it for each usage case.
The influence of pulse delay Δt between the driving pulse and the prepulse was measured using the setup presented in Fig. 1(b). The experimentally measured dependence of the generated X-rays on the interpulse pulse delay is shown in Fig. 4. The amplitude ratio of the prepulse to the main pulse was fixed at 1/10 (7.2/72.2 J/cm2 fluence on the target surface), with 0.55 mJ pulse packet energy. The X-ray yield exhibits a clear maximum with respect to interpulse delay, which is observed at about 125 ps. With the further delay of the main pulse, the yield gradually decreases again, and the observed decrease is steeper at 25 kHz repetition rate compared to 100 kHz (repetition rate was changed using pulse picker). For 100 kHz, the optimal prepulse delay Δt range is quite broad (from 100 to 500 ps). We note that the entire explored delay range of 10 ps to 1600 ps leads to a marked increase of the generated X-ray radiation. In contrast, when tested in ns burst mode with prepulse timed 13.15 ns before the main pulse, no prepulse-induced enhancement could be observed, indicating that the prepulse effect effectively vanishes within this time period.
Fig. 4. X-ray emission in 6–40 keV range dependence on the prepulse to driving pulse delay at 100 kHz and 25 kHz pulse repetition rates. The energy of the pulse packet is 0.55 mJ, and the prepulse and driving pulse amplitude ratio is 1/10 (7.2/72.2 J/cm2 fluence on the target surface), respectively. Every data point was integrated for 500 000 laser pulses.
The interpretations of the observed dependence of LPP X-ray generation efficiency on prepulse amplitude and have been offered by several authors [38–40,47]. By controlling the delay time and amplitude of the prepulse, we can control the plasma density scale length, i.e., the dimensions of plasma volume with approximately constant parameters. This, in turn, controls light absorption in an expanding preplasma. As the prepulse to the driving pulse delay and the beam diameter is fixed, there exists an optimal prepulse amplitude for the most efficient X-ray generation. If the amplitude of the prepulse and the beam diameter is fixed, there exists an optimal delay time Δt. Moreover, it is shown that at relatively small scale length (compared to the wavelength of laser pulse), dense electron plasma effectively acts as a mirror reflecting the light of the driving pulse [48,49]. Thus the right prepulse amplitude and delay are crucial for effective driving pulse absorption. In addition, the electron collision model also suggested that the electron mean free path is an important parameter in the generation of ultrafast pulsed X-rays in any ambient condition: X-ray emission increases with increasing electron free path until it reaches a plateau [50].
We also observed an interesting tendency that total X-ray photon count produced by a fixed number of laser pulses increased with increasing laser repetition rate (in the range of 5–100 kHz) when the pulse packet energy was kept fixed (see Fig. 5(a)). This increase stopped and was reversed slightly before the subsequent pulse-affected areas on the target began to overlap. It is important to mention that the repetition rate was changed using pulse picker. Thus, no significant change in beam parameters is expected. Presumably, the insufficient target refreshment results in the decrease of X-ray flux at higher laser repetition rates [27]. To have higher overlap and to see it more clearly, the target rotation speed was reduced to 300 RPM during the measurement (presented in Fig. 5). It can be seen that the relative shape of the X-ray spectrum also changes (see Fig. 5(b)), and relatively more photons are concentrated in Kα line at a higher repetition rate (see in Fig. 5(a) the red line). Moreover, an increase in the repetition rate leads to a slightly increased optimal delay time range (see Fig. 4). These observations indicate that plasma dynamics are influenced by the aggregated effect of multiple pulses on the target.
Fig. 5. (a) X-ray emission in 6–40 keV range (black line) and percentage of photons at Kα line (500 eV bandwidth) in this range (red line) dependence on laser repetition rate when prepulse and driving pulse amplitude ratio is 1/10 (11.8/118 J/cm2 fluence on the target surface) respectively, pulse packet energy is 0.9 mJ, and the target rotation speed is ∼ 300 RPM. The top x-axis represents calculated relative pulse-affected area separation (compared to laser spot size) between consequent pulses. (b) The acquired spectrum of X-ray radiation at different repetition rates.
To investigate this phenomenon more deeply, we measured the X-ray emission dependencies on the separation of pulse packets on the microsecond time scale. The microsecond burst of the pulse packets was generated with the fixed average power of 4.5 W (see Fig. 6(a)). Different irradiation protocols were investigated by grouping prepulse-pulse pairs into microsecond bursts containing N = 1,2,3… such pairs separated by 10 µs. The average 4.5 W power was kept the same by correspondingly increasing the waiting time between microsecond bursts, as shown in Fig. 6(a). For example, N = 1 corresponds to single prepulse-pulse pair arriving every 200 µs, whereas N = 4 corresponds to a packet of four prepulse-pulse pairs arriving with 10 µs time intervals and then no radiation for the remaining portion of 800 µs period. In both cases the total energy received over the longer period is the same, and so is the average power. The energy of prepulse-pulse pair was fixed at 0.9 mJ, and the disk rotation speed was set at 800 RPM.
Fig. 6. (a) Schematic depiction of microsecond/picosecond pulse burst sequences, at which X-ray emission data was collected. (b) X-ray emission in 6–40 keV range dependence on the number of pulse pairs in a microsecond burst while retaining the same average beam power (columns) and the calculated X-ray emission value generated from subsequent pulse packets in microsecond burst Ip > 1 (red squares).
Figure 6(b) shows experimental results of X-ray emission increase because of the microsecond burst packet shaping with 10 µs between the pulse packets. Even though the average power is the same, the plasma dynamics favor not only the prepulse and the main pulse picosecond time scale delay but also the pulse packets deliberately shaped in the microsecond time scale. In addition, a simple numerical calculation was done (see Eq. (2) and Eq. (3)). The registered X-ray emission could be proportionally divided in the contribution of the first (${p_1}$) and the subsequent ($p > 1$) pulse packets:
The ideas for a qualitative explanation of this observation may be gleaned from the studies of femtosecond-pulse-induced ablation plumes on a microsecond time scale [51,52]. Typically, femtosecond LPP plumes expand with a much stronger forward bias in directions normal to the target surface than in parallel. Ablation plume typically contains nanoparticles and atoms of ablated material [53]. When a target rotation speed is high enough to greatly separate areas of consequent pulse packets, or the repetition rate is very low (for our experiment ∼ 300 RPM and <5 kHz, see Fig. 5(a)), subsequent pulse packets do not “see” any ablated material nanoparticles and atoms cloud and interacts only with a fresh surface. In contrast, if the target is spinning too slowly or the repetition rate is too high (for our experiment ∼ 300 RPM and > 40 kHz, see Fig. 5(a)), some part of subsequent pulse packet energy is scattered and absorbed by an expanded ablation plume relatively far away from the target, and X-ray emission drops.
Presumably, a favorable condition exists (for our experiment ∼ 300 RPM and 5 - 40 kHz, see Fig. 5(a)) when the ablated material cloud is located near the target surface, and relatively a small part of the subsequent pulse packet is scattered. Intuitively, material plume increases an effective surface area and might increase overall laser absorption. On the other hand, the accelerated electrons interact not only with a few microns of the target surface but also with earlier ablated particles. From Fig. 5(b), it can be seen that the X-ray photon spectra enlarge to higher energies and the relative proportion of characteristic K-radiation increases with increasing repetition rate. It indicates that hot electron temperature also increases [1,15]. The hot electron temperature can be derived from Fig. 7 by fitting a Maxwellian electron distribution function [50]. Figure 7 shows the spectra of bremsstrahlung emission in the range of 8–20 keV, which was calculated using Eq. (1) and the data presented in Fig. 5(b). The calculated hot electron temperature at 5 kHz is 3.44 keV, and it rises to 4.33 keV at 33.3 kHz and slightly decreases to 4.05 keV at 100 kHz. It shows that electron acceleration depends on repetition rate, which might be a consequence of changes in driving pulse absorption or electron free path [50]. However, this speculative explanation should be verified with further experiments by correlating ablation plume dynamical changes with changes in X-ray emission.
Fig. 7. Spectra of high-energy bremsstrahlung emission in the range 8–20 keV at different repetition rates when prepulse and driving pulse amplitude ratio is 1/10 respectively, pulse packet energy is 0.9 mJ, and the target rotation speed is ∼ 300 RPM. The X-ray intensity was fitted by a Maxwellian electron distribution and the hot electron temperature was derived.
In summary, the interplay between ablated material plume scattering and some enhancement mechanism will determine the overall X-ray yield increase. It is also worth mentioning that a strong influence of the temporal and spatial inter-pulse distance on X-ray emission has already been reported in material processing [44].
3.2 X-ray photon flux measurement
The X-ray flux (into 4π solid angle) dependence on the driving pulse energy in both a single pulse and a pair of prepulse and the driving pulse cases are shown in Fig. 8(a). The measurements were performed using Fe, Cu, and Al target materials, and focusing conditions were optimized before every measurement point. The optimum prepulse to the driving pulse energy ratio for all target materials was around 1/10. The time separation between the pulses in a pulse packet was 440 ps, and the repetition rate was 100 kHz. The flux values were calculated using Eq. (1) at the target surface.
Fig. 8. (a) X-ray flux in 6–40 keV range and into 4π solid angle and (b) calculated spectral photon flux into 4π angle for different target materials at 100 kHz repetition rate. The driving pulse energy without prepulse is EDR = 0.9 mJ, the driving pulse energy with prepulse is EDR = 0.82 mJ, and the prepulse energy is EPP = 0.08 mJ.
In the prepulse case, Fig. 8(a) shows the highest registered X-ray photon flux (in 6–40 keV range), reaching 8 × 1011 photons/s for Fe target, 2.8 × 1011 photons/s for Cu and 0.7 × 1011 photons/s for Al target, with the calculated energy conversion efficiency of 1 × 10−5, 4.3 × 10−6, 1.3 × 10−6 respectively. The highest X-ray photon flux in the single-pulse regime was nearly 300 times lower, 2.8 × 109 photons/s with 3.5 × 10−8 conversion efficiency.
The corresponding spectral photon fluxes are presented in Fig. 8(b). The spectral photon flux of the presented source is more than one order of magnitude higher compared to existing experimental stations [3,54] and has around one order lower pulse energy.
Table 1 represents the highest registered photon flux into 4π solid angle and efficiency for different target materials in the 6–40 keV range and at the Kα line. The Kα flux is calculated by integrating a spectral area of 500 eV centered on the characteristic line peak (Fe: ∼ 6.4 keV, Cu: ∼ 8.04 keV). It can be seen that 70% (5.5 × 1011 photons/s) of X-ray photons and 60% (6.14 × 10−6) of total energy is located at Fe Kα line and 43% (1.21 × 1011 photons/s) of X-ray photons and 40% (1.71 × 10−6) of total energy is located at Cu Kα line in the spectral range of 6–40 keV. These calculated flux values are significantly larger than the ones found in previous works performed in the ambient air environment, using metal targets and comparable pulse intensities [27]. The reason for higher observed flux values is the combination of high average laser power (90 W) of the laser pulses and the deliberately shaped prepulse. Moreover, these values are even comparable with a novel table-top hard X-ray source driven by femtosecond mid-infrared pulses in vacuum generating 1.5 × 1012 photons/s at Cu Kα [17].
Table 1. The highest detected photon fluxes into 4π solid angle for different materials in 6–40 keV range and at the Kα line of 500 eV bandwidth.
It is important to note that single pulse conversion efficiency measured in our work (10−8) is almost an order of magnitude lower than those obtained by other authors under similar conditions [27]. We attribute this to sub-optimal target disk balancing, which led to variations of disk position with respect to the laser focus [14] and non-ideal debris removal (simple shop vacuum suction), causing some X-ray reabsorption by ablated material. Even higher X-ray flux values may be expected if these issues are addressed.
In this work, the X-ray source size is not characterized and will be done in future experiments using a more stable target setup. However, the size of the radiating plasma on the laser intensity typically is just a few times larger than the laser focal spot and might be expected to be 2–9 times larger [11]. It means that in our case, the X-ray source size might be in the 50–225 µm (FWHM) range, and the expected average brightness of Fe Kα and Cu Kα line is in 5 × 109–2 × 108 ph/(s mm2 mrad2 (0.1% BW)) and 1.5 × 109–6 × 107 ph/(s mm2 mrad2 (0.1% BW)) range respectively.
Data presented in Fig. 8(a) clearly shows that the X-ray emission and conversion efficiency start to saturate at the driving intensities of approximately 2 × 1014 W/cm2. At the same time, the X-ray emission at the saturation intensity for pulse pairs is way above that for single pulses. The reason for this is that not only is the fraction of pulse energy allocated to prepulse used productively (i.e. without clamping), it also alters the target into the pre-plasma state, where the energy of the main pulse can be converted into X-rays much more efficiently. There is a consensus in the literature that X-ray flux follows a power law with respect to the driving light intensity $\Phi \; \sim \; {I^\alpha }$, where I represent laser intensity and $\alpha $ represents scaling factor, which was estimated to be ≈ 2 at 1014–15 W/cm2 intensity [27,55] and even larger (>2) in case of larger intensities [19]. The most obvious way of increasing the intensity is the operation in the atmosphere of noble gas with higher ionization energy (He, Ne). If the scaling law holds for pulse pair packet operation, such modification of the setup will allow expecting 1012 - 1013 photon flux into 4π angle with 10−5–10−4 conversion efficiency.
The ultrashort time duration of the X-ray pulses is one of the essential features of the LPP X-ray source and it is might be comparable with that of the laser pulse producing the plasma [1]. However, it is important to mention some tradeoffs between enhancing the X-ray emission and maintaining a short pulse duration. It is shown that with increasing foil target thickness X-ray generation efficiency increases; however, the X-ray pulse duration also increases [56]; this also happens with increasing laser intensity when the foil thickness is fixed [57]. Nanostructured target X-ray yield enhancement could also lead to X-ray pulse broadening [32]. It is shown that both the X-ray emission and the pulse duration increased with an increase in the scale length of the prepulse-created plasma [40]. In particular, the rise time of the X-ray emission is mainly determined by the characteristic time needed to heat the plasma, while the time scale of plasma cooling sets the fall time of the emission. The higher electron temperature, which is related to laser intensity or lower plasma heat conductivity, which is related to plasma scale length, implies longer afterglow emission. In summary, intensity X-ray flux scaling or prepulse enhancement both increases X-ray pulse duration. Thus we expect that the presented X-ray source pulse duration should be comparable to a few orders higher intensity vacuum-based systems [3,54,58] and for >3 keV photon energies should be less than 10 picoseconds. However, this needs to be confirmed with the further X-ray pulse duration measurements.
3.3 Lifetime of the X-ray source
Another (unintended) benefit of using pulse pair for X-ray generation turned out to be the lifetime of the target under high-intensity femtosecond laser irradiation. It turned out that the pulse pair produces a significantly milder ablation of a target, and the stable generation can be maintained for a more extended period of time. The results of the long-term stability test of the X-ray source are shown in Fig. 9 in two operating regimes: 1) a single 0.9 mJ pulse is focused on the sample; 2) a prepulse and driving pulse pair are formed with 440 ps delay between the pulses in a pulse packet and 10% prepulse to the driving pulse energy ratio (FPP = 11.8 J/cm2, FDR = 118 J/cm2 fluence on the target surface), the packet energy is 0.9 mJ. Each data point was averaged over a time interval of 30 s or 3 million pulses. In both cases, the iron target is irradiated with similar >400 TW/cm2 intensity. The target was rotated at 800 RPM rotational speed and slowly translated back and forth along the disk radius. A single translation took 30 seconds; therefore, during 60 min measurement target was translated 120 times through the same target area. As can be seen from Fig. 9(a), the two pulse regime not only increases X-ray emission by two orders but also leads to significantly better short-term stability and substantially increases a target lifetime. In addition, the collected X-ray spectra are shown in Fig. 9(b).
Fig. 9. Long-term performance of the X-ray source presenting the time dependence of the generated (a) X-ray intensity and (b) X-ray spectra in the pulse packet and the single pulse operation.
Poor short-term stability in the single pulse case is linked to the pronounced ablation of the material and change of the surface properties, such as roughness and an effective area, which governs the light absorption properties and overall X-ray emission [59,60]. In a single pulse case, the generated X-ray intensity increases at first (favorable modification of the target) and decreases later (detrimental surface modification). In the two-pulse case, the prepulse generates plasma, which efficiently absorbs the driving pulse energy and effectively shields the target [53]. It would be tempting to attribute the entire stability increase to the amount of ablated material. However, the removed target mass only differs by the factor of 1.5 for pulse pair compared to single pulses (see Table 2).
Table 2. Ablated mass of a target material per 100k laser pulses when a target is irradiated with pulse packets or single pulses.
The complementing explanation might be that the prepulse not only shields a target but also produces different crater surface geometry, keeping the surface smooth for longer. It is shown that compared to the single pulse regime, a reduction of the surface roughness by a factor of four could be accomplished depending on the number of pulses per burst [61]. This assumption was experimentally tested by recreating spinning disk conditions with a galvanoscanner and scanning the iron surface multiple times. Comparing 0.4 mJ single pulse target ablation with 0.4 mJ pulse packet operation, when prepulse and driving pulse amplitude ratio was 1/10 (5.2/52.5 J/cm2 fluence on the target surface) respectively, it can be seen that in the single pulse case the roughness of the surface increases gradually when in the pulse packet case roughness almost do not change (see Fig. 10(a)). The roughness difference is more than two times after 250 scans (it corresponds to 2 hours of spinning disk operation). Moreover, surfaces scanned with the single pulse have randomly located craters that are several times smaller than beam spot size, and there is no indication of craters in the pulse packet case (see Fig. 10(b–c)). These craters might highly affect pulse to pulse stability in the single pulse case.
Fig. 10. (a) Surface roughness dependence on the number scans recreating spinning disk scanning conditions with galvanoscanner in single pulse and pulse packet case. The images of the iron surface after 250 scans in (b) pulse packet (c) single pulse case.
3.4 Proof-of-principle applications: imaging and spectroscopy
We performed two demonstration experiments employing the femtosecond X-ray source for imaging and spectroscopy. The first one was X-ray absorption imaging using a CMOS imaging sensor (Hamamatsu C9732DK-11). The sensor was placed ∼ 30 cm from the Fe X-ray source (100 kHz repetition rate) and the acquisition time for each image was 2 seconds. The mock-up sample consisted of some iron washers and a piece of glass between two sheets of paper sealed with electrical tape (Fig. 11(a–b)). The sample was placed immediately before the detector. The acquired X-ray images are presented in Fig. 11(c-e), where the same image was recorded with aluminum sheets of different thicknesses were placed between the source and the sample. The metal and glass parts in the sample are clearly discerned with 0.5, and metal parts are discerned even with 2 mm Al shielding (see Fig. 11(d–e)).
Fig. 11. (a) Metal and glass parts in the sample and (b) the final sample covered with tape. (c) X-ray image of the sample at 30 cm from the target. X-ray image with (d) 0.5 mm and (e) 2 mm aluminum plate before the sample.
The second demonstration was intended to explore the X-ray source as a candidate for XAS applications. It must be noted, that for time-resolved XAS experiments, a few eV spectral resolution would be preferable [2,4]. Such resolution might be accomplished using a dispersive spectrograph with microcalorimeter array detectors [3,62,63] or a dispersion-less spectrograph using deep depletion X-ray CCD [64,65]. Therefore, the presented results obtained by the SDD spectrometer should only be regarded as a mock-up or demonstration. In addition to the relatively poor spectral resolution (∼140 eV), the second limitation of this spectrometer was related to the input rate of the SSD. Under pulsed illumination, it is necessary to keep the counting rate below 1% of the applied laser pulse repetition rate to reduce the photon pile-up phenomena significantly. It means that operating the laser at 100 kHz, only 1000 photons/s can be registered, considerably increasing the measurement time. This limitation becomes even more severe in the case of 1 kHz repetition rate lasers widely used in ultrafast research. We opted to increase the counting rate by increasing the laser repetition rate to 2 MHz (EPP = 5 µJ and EDR = 40 µJ), which enabled us to register 20000 photons/s. To compensate for a massive drop in the X-ray generation efficiency, tighter beam focusing (f = 75 mm) was used, which allowed 2 times smaller beam diameter and ∼ 80 TW/cm2 intensity onto aluminum target surface at 40 µJ pulse energy. This corresponds to FPP = 2.1 J/cm2, FDR = 21 J/cm2 fluence on the target surface. In addition, the target rotational speed was increased to ∼ 1600 RPM, and the spectrometer was placed 60 cm from the target. The calculated difference of spectral photon flux at 100 kHz and 2 MHz is presented in Fig. 12. As can be seen, the total flux is more than 10 times lower; however, increased photon-counting rate allowed us to have ∼20 times shorter measurement time (compared to the most efficient 100 kHz operation) for an optimal signal-to-noise ratio.
Fig. 12. Calculated spectral photon flux into 4π angle, using Al target at 100 kHz and 2 MHz repetition rate.
Two X-ray spectrometers were employed side-by-side for XAS two-channel measurements, one of which registered the reference spectrum and another registered the spectrum of photons transmitted through a sample. By dividing the two signals, the transmittance spectrum of the sample was calculated. To obtain solid samples with controllable X-ray transmission, we used 0.18 mm thick microscope coverslips coated with Cu layers of different thicknesses using a rotary pumped coater (Quorum Q150R S). The transmittance spectrum dependence on the Cu sputtering time in the coater is presented in Fig. 13(a). Each transmittance spectrum was integrated for 3 minutes and averaged 3 times, taking around 10 minutes for one measurement in total. The apparent changes in the sample transmittance spectrum can be seen after 15 min of sputtering, and this signal is comparable to the theoretical transmittance spectrum of 350 nm Cu. The recorded spectra of reference and sample spectrometers are presented in Fig. 13(b), where 5% in the transmission change can be discerned. This would also be the sensitivity of the time-resolved pump-probe experiment, where the signal is constructed by comparing pumped and unpumped samples. In addition, the discussed source flux is comparable with already demonstrated time-resolved XAS experiments using LPP X-rays [3,54]. However, X-ray pulse durations should be measured by further experiments to disclose the true source potential for such applications.
Fig. 13. (a) Sample transmittance spectrum dependence on Cu sputtering time of the sample. The thick black and red line represent the calculated transmittance of 350 nm and 750 nm Cu layers. (b) Recorded spectra of reference and sample spectrometers after 50 min of Cu sputtering.
4. Conclusions
We have demonstrated and characterized a high-efficiency compact and robust laser-produced plasma X-ray source driven by double-pulsed femtosecond laser pulses at 100 kHz repetition rate in an ambient air environment. The deliberately shaped prepulse and driving pulse pair increases X-ray emission by more than two orders of magnitude, significantly expands the lifetime of a rotating disk target and ensures better short-term X-ray emission stability compared to the same energy single pulse operation. The obtained photon flux, and conversion efficiency are comparable to multi-milijoule pulsed laser systems operating in a vacuum environment. Its simplicity allowed us to demonstrate proof-of-principle applications in imaging and spectroscopy. The results allow us to expect the successful application of the presented method for femtosecond and picosecond time-resolved X-ray experiments.
Funding
Lietuvos Mokslo Taryba (Research Council of Lithuania) (09.3.3-LMT-K-712-0014, 09.3.3-LMT-K-712-19-0014).
Acknowledgments
L.R., S.B. and M.V. acknowledge funding from European Social Fund according to the activity ‘Improvement of researchers’ qualification by implementing world-class R&D projects’ of Measure No. 574 09.3.3-LMT-K-712, grant No. 09.3.3-LMT-K-712-0014. V.B. acknowledges funding from European Social Fund (project No 09.3.3-LMT-K-712-19-0014) under grant agreement with the Research Council of Lithuania (LMTLT).
Disclosures
The authors hereby disclose the following conflicts of interest: Lukas Rimkus, Simas Butkus and Mikas Vengris are paid employees of Light Conversion, the manufacturer of the laser used as a driving source for X-ray generation. The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. D. Giulietti and L. A. Gizzi, “X-ray emission from laser-produced plasmas,” Riv. Nuovo Cim. 21(10), 1–93 (1998). [CrossRef]
2. M. Chergui and A. H. Zewail, “Electron and X-Ray Methods of Ultrafast Structural Dynamics: Advances and Applications,” Chemphyschem 10(1), 28–43 (2009). [CrossRef]
3. L. Miaja-Avila, G. C. O’Neil, J. Uhlig, C. L. Cromer, M. L. Dowell, R. Jimenez, A. S. Hoover, K. L. Silverman, and J. N. Ullom, “Laser plasma x-ray source for ultrafast time-resolved x-ray absorption spectroscopy,” Struct. Dyn. 2, 024301 (2015). [CrossRef]
4. C. Bressler and M. Chergui, “Ultrafast X-ray absorption spectroscopy,” Chem. Rev. 104(4), 1781–1812 (2004). [CrossRef]
5. L. X. Chen, “Probing transient molecular structures in photochemical processes using laser-initiated time-resolved X-ray absorption spectroscopy,” in Annual Review of Physical Chemistry (Annual Reviews, Palo Alto, 2005), pp. 221–254.
6. M. Afshari, P. Krumey, D. Menn, M. Nicoul, F. Brinks, A. Tarasevitch, and K. Sokolowski-Tinten, “Time-resolved diffraction with an optimized short pulse laser plasma X-ray source,” Struct. Dyn.-US 7, 014301 (2020). [CrossRef]
7. F. Zamponi, Z. Ansari, M. Woerner, and T. Elsaesser, “Femtosecond powder diffraction with a laser-driven hard X-ray source,” Opt. Express 18(2), 947–961 (2010). [CrossRef]
8. C. Rose-Petruck, R. Jimenez, T. Guo, A. Cavalleri, C. W. Siders, F. Raksi, J. A. Squier, B. C. Walker, K. R. Wilson, and C. P. J. Barty, “Picosecond-milliangstrom lattice dynamics measured by ultrafast X-ray diffraction,” Nature 398(6725), 310–312 (1999). [CrossRef]
9. P. Chen, I. V. Tomov, and P. M. Rentzepis, “Time resolved heat propagation in a gold crystal by means of picosecond x-ray diffraction,” J. Chem. Phys. 104(24), 10001–10007 (1996). [CrossRef]
10. A. S. Martynenko, S. A. Pikuz, I. Y. Skobelev, S. N. Ryazantsev, C. D. Baird, N. Booth, L. N. K. Dohl, P. Durey, A. Y. Faenov, D. Farley, R. Kodama, K. Lancaster, P. McKenna, C. D. Murphy, C. Spindloe, T. A. Pikuz, and N. Woolsey, “Optimization of a laser plasma-based x-ray source according to WDM absorption spectroscopy requirements,” Matter Radiat. Extrem. 6(1), 014405 (2021). [CrossRef]
11. R. Toth, S. Fourmaux, T. Ozaki, M. Servol, J. C. Kieffer, R. E. Kincaid, and A. Krol, “Evaluation of ultrafast laser-based hard x-ray sources for phase-contrast imaging,” Phys. Plasmas 14(5), 053506 (2007). [CrossRef]
12. A. Krol, J. C. Kieffer, J. Nees, L. Chen, R. Toth, B. Hou, R. E. Kincaid, I. L. Coman, E. D. Lipson, and G. Mourou, “Ultrafast laser-based micro-CT system for small animal imaging,” in Medical Imaging 2004 Conference, (Spie-Int Soc Optical Engineering, 2004), 265–271.
13. M. Gambari, R. Clady, A. Stolidi, O. Uteza, M. Sentis, and A. Ferre, “Exploring phase contrast imaging with a laser-based K-alpha x-ray source up to relativistic laser intensity,” Sci. Rep. 10(1), 6766 (2020). [CrossRef]
14. L. Martin, J. Benlliure, D. Cortina-Gil, J. Penas, and C. Ruiz, “Improved stability of a compact vacuum-free laser-plasma X-ray source,” High Power Laser Sci. Eng. 8, e18 (2020). [CrossRef]
15. P. Gibbon and E. Forster, “Short-pulse laser-plasma interactions,” Plasma Phys. Control. Fusion 38(6), 769–793 (1996). [CrossRef]
16. M. M. Murnane, H. C. Kapteyn, M. D. Rosen, and R. W. Falcone, “ULTRAFAST X-RAY PULSES FROM LASER-PRODUCED PLASMAS,” Science 251(4993), 531–536 (1991). [CrossRef]
17. A. Koç, C. Hauf, M. Woerner, L. von grafensteın, D. Ueberschaer, M. Bock, U. Griebner, and T. Elsaesser, “Compact high-flux hard X-ray source driven by femtosecond mid-infrared pulses at a 1 kHz repetition rate,” Opt. Lett. 46(2), 210–213 (2021). [CrossRef]
18. M. Yoshida, Y. Fujimoto, Y. Hironaka, K. G. Nakamura, K. Kondo, M. Ohtani, and H. Tsunemi, “Generation of picosecond hard x rays by tera watt laser focusing on a copper target,” Appl. Phys. Lett. 73(17), 2393–2395 (1998). [CrossRef]
19. V. Arora, P. A. Naik, J. A. Chakera, S. Bagchi, M. Tayyab, and P. D. Gupta, “Study of 1-8 keV K-alpha x-ray emission from high intensity femtosecond laser produced plasma,” AIP Adv. 4(4), 047106 (2014). [CrossRef]
20. A. Y. Faenov, J. Colgan, S. B. Hansen, A. Zhidkov, T. A. Pikuz, M. Nishiuchi, S. A. Pikuz, I. Y. Skobelev, J. Abdallah, H. Sakaki, A. Sagisaka, A. S. Pirozhkov, K. Ogura, Y. Fukuda, M. Kanasaki, N. Hasegawa, M. Nishikino, M. Kando, Y. Watanabe, T. Kawachi, S. Masuda, T. Hosokai, R. Kodama, and K. Kondo, “Nonlinear increase of X-ray intensities from thin foils irradiated with a 200 TW femtosecond laser,” Scientific Reports 5(2015).
21. A. A. Garmatina, B. G. Bravy, F. V. Potemkin, M. M. Nazarov, and V. M. Gordienko, “Intensity clamping and controlled efficiency of X-ray generation under femtosecond laser interaction with nanostructured target in air and helium,” J. Phys.: Conf. Ser. 1692(1), 012004 (2020). [CrossRef]
22. B. X. Hou, J. Easter, A. Mordovanakis, K. Krushelnick, and J. A. Nees, “Vacuum-free x-ray source based on ultrashort laser irradiation of solids,” Opt. Express 16(22), 17695–17705 (2008). [CrossRef]
23. S. Amano, Y. Inaoka, H. Hiraishi, S. Miyamoto, and T. Mochizuki, “Laser-plasma debris from a rotating cryogenic-solid-Xe target,” Rev. Sci. Instrum. 81(2), 023104 (2010). [CrossRef]
24. S. Amano, T. Inoue, and T. Harada, “Plasma debris sputter resistant x-ray mirror,” Appl. Opt. 52(16), 3845–3848 (2013). [CrossRef]
25. R. Alnaimi, “Development of a low-debris laser driven tape drive soft x-ray source,” J. Instrum. 12(06), P06001 (2017). [CrossRef]
26. M. Hagedorn, J. Kutzner, G. Tsilimis, and H. Zacharias, “High-repetition-rate hard X-ray generation with sub-millijoule femtosecond laser pulses,” Appl. Phys. B 77(1), 49–57 (2003). [CrossRef]
27. A. Baguckis, A. Plukis, J. Reklaitis, V. Remeikis, L. Giniunas, and M. Vengris, “Generation of plasma X-ray sources via high repetition rate femtosecond laser pulses,” Appl. Phys. B 123(12), 290 (2017). [CrossRef]
28. C. Bai, W. L. Tian, G. Y. Wang, L. Zhen, R. Xu, D. C. Zhang, Z. H. Wang, J. F. Zhu, and Z. Y. Wei, “Progress on Yb-Doped All-Solid-State Femtosecond Laser Amplifier with High Repetition Rate,” Chin. J. Lasers 48, 16 (2021).
29. D. Rolles, “X-RAY OPTICS Highly efficient nanoscale X-ray sources,” Nat. Photonics 12(2), 62–63 (2018). [CrossRef]
30. T. Ebert, N. W. Neumann, L. N. K. Dohl, J. Jarrett, C. Baird, R. Heathcote, M. Hesse, A. Hughes, P. McKenna, D. Neely, D. Rusby, G. Schaumann, C. Spindloe, A. Tebartz, N. Woolsey, and M. Roth, “Enhanced brightness of a laser-driven x-ray and particle source by microstructured surfaces of silicon targets,” Phys. Plasmas 27(4), 043106 (2020). [CrossRef]
31. U. Chakravarty, V. Arora, P. A. Naik, J. A. Chakera, H. Srivastava, A. Srivastava, G. D. Varma, S. R. Kumbhare, and P. D. Gupta, “Enhancement of Kα emission through efficient hot electron generation in carbon nanotubes on intense laser pulse irradiation,” J. Appl. Phys. 112(5), 053301 (2012). [CrossRef]
32. T. Nishikawa, H. Nakano, K. Oguri, N. Uesugi, K. Nishio, and H. Masuda, “Nanohole-array size dependence of soft x-ray generation enhancement from femtosecond-laser-produced plasma,” J. Appl. Phys. 96(12), 7537–7543 (2004). [CrossRef]
33. O. Klimo, J. Psikal, J. Limpouch, J. Proska, T. Ceccotti, V. Floquet, and S. Kawata, “Simulations of short pulses laser interaction with targets having a submicron surface structure - energy absorption and ion acceleration,” in Conference on Laser Acceleration of Electrons, Protons, and Ions and Medical Applications of Laser-Generated Secondary Sources of Radiation and Particles, (Spie-Int Soc Optical Engineering, 2011).
34. A. Andreev, J. Imgrunt, V. Braun, I. Dittmar, and U. Teubner, “Towards optimization of femtosecond laser pulse nanostructuring of targets for high-intensity laser experiments in vacuum,” Appl. Phys. A 127(7), 564 (2021). [CrossRef]
35. T. Nishikawa, H. Nakano, H. Ahn, N. Uesugi, and T. Serikawa, “X-ray generation enhancement from a laser-produced plasma with a porous silicon target,” Appl. Phys. Lett. 70(13), 1653–1655 (1997). [CrossRef]
36. D. Kühlke, U. Herpers, and D. V. D. Linde, “Soft x-ray emission from subpicosecond laser-produced plasmas,” Appl. Phys. Lett. 50(25), 1785–1787 (1987). [CrossRef]
37. R. Lokasani, G. Arai, Y. Kondo, H. Hara, T.-H. Dinh, T. Ejima, T. Hatano, W. Jiang, T. Makimura, B. Li, P. Dunne, G. O’Sullivan, T. Higashiguchi, and J. Limpouch, “Soft X-ray emission from molybdenum plasmas generated by dual laser pulses,” Appl. Phys. Lett. 109(19), 194103 (2016). [CrossRef]
38. W. Lu, M. Nicoul, U. Shymanovich, A. Tarasevitch, P. Zhou, K. Sokolowski-Tinten, D. von der Linde, M. Mašek, P. Gibbon, and U. Teubner, “Optimized K α x-ray flashes from femtosecond-laser-irradiated foils,” Phys. Rev. E 80(2), 026404 (2009). [CrossRef]
39. P. Gibbon, M. Masek, U. Teubner, W. Lu, M. Nicoul, U. Shymanovich, A. Tarasevitch, P. Zhou, K. Sokolowski-Tinten, and D. von der Linde, “Modelling and optimisation of fs laser-produced K (alpha) sources,” Appl. Phys. A 96(1), 23–31 (2009). [CrossRef]
40. H. Nakano, T. Nishikawa, H. Ahn, and N. Uesugi, “Temporal evolution of soft x-ray pulse emitted from aluminum plasma produced by a pair of Ti:sapphire laser pulses,” Appl. Phys. Lett. 69(20), 2992–2994 (1996). [CrossRef]
41. J. Uhlig, C. G. Wahlstrom, M. Walczak, V. Sundstrom, and W. Fullagar, “Laser generated 300 keV electron beams from water,” Laser Part. Beams 29(4), 415–424 (2011). [CrossRef]
42. M. Barkauskas, K. Neimontas, and V. Butkus, “Device and method for generation of high repetition rate laser pulse bursts,” (Google Patents, 2020).
43. Y. Hironaka, K. G. Nakamura, and K.-I. Kondo, “Angular distribution of x-ray emission from a copper target irradiated with a femtosecond laser,” Appl. Phys. Lett. 77(25), 4110–4111 (2000). [CrossRef]
44. J. Schille, S. Kraft, T. Pflug, C. Scholz, M. Clair, A. Horn, and U. Loeschner, “Study on X-ray Emission Using Ultrashort Pulsed Lasers in Materials Processing,” Materials 14(16), 4537 (2021). [CrossRef]
45. D. Metzner, M. Olbrich, P. Lickschat, A. Horn, and S. Weissmantel, “X-ray generation by laser ablation using MHz to GHz pulse bursts,” J. Laser Appl. 33(3), 032014 (2021). [CrossRef]
46. A. Couairon and A. Mysyrowicz, “Self-focusing and filamentation of femtosecond pulses in air and condensed matter: simulations and experiments,” Self-focusing: Past and Present, 297–322 (2009).
47. S. Bastiani, A. Rousse, J. P. Geindre, P. Audebert, C. Quoix, G. Hamoniaux, A. Antonetti, and J. C. Gauthier, “Experimental study of the interaction of subpicosecond laser pulses with solid targets of varying initial scale lengths,” Phys. Rev. E 56(6), 7179–7185 (1997). [CrossRef]
48. G. G. Scott, G. F. H. Indorf, M. A. Ennen, P. Forestier-Colleoni, S. J. Hawkes, L. Scaife, M. Sedov, D. R. Symes, C. Thornton, F. Beg, T. Ma, P. McKenna, A. A. Andreev, U. Teubner, and D. Neely, “Kinematics of femtosecond laser-generated plasma expansion: Determination of sub-micron density gradient and collisionality evolution of over-critical laser plasmas,” Phys. Plasmas 28(9), 093109 (2021). [CrossRef]
49. G. G. Scott, V. Bagnoud, C. Brabetz, R. J. Clarke, J. S. Green, R. I. Heathcote, H. W. Powell, B. Zielbauer, T. D. Arber, P. McKenna, and D. Neely, “Optimization of plasma mirror reflectivity and optical quality using double laser pulses,” New J. Phys. 17(3), 033027 (2015). [CrossRef]
50. M. Hada and J. Matsuo, “Effects of ambient pressure on Cu Kα X-ray radiation with millijoule and high-repetition-rate femtosecond laser,” Appl. Phys. B 99(1-2), 173–179 (2010). [CrossRef]
51. T. Donnelly, G. O’Connell, and J. G. Lunney, “Metal Nanoparticle Film Deposition by Femtosecond Laser Ablation at Atmospheric Pressure,” Nanomaterials 10(11), 2118 (2020). [CrossRef]
52. B. Verhoff, S. S. Harilal, J. R. Freeman, P. K. Diwakar, and A. Hassanein, “Dynamics of femto- and nanosecond laser ablation plumes investigated using optical emission spectroscopy,” J. Appl. Phys. 112(9), 093303 (2012). [CrossRef]
53. S. Noel and J. Hermann, “Reducing nanoparticles in metal ablation plumes produced by two delayed short laser pulses,” Appl. Phys. Lett. 94(5), 053120 (2009). [CrossRef]
54. F. Dorchies, N. Fedorov, and L. Lecherbourg, “Experimental station for laser-based picosecond time-resolved x-ray absorption near-edge spectroscopy,” Rev. Sci. Instrum. 86(7), 073106 (2015). [CrossRef]
55. J. Badziak, J. Makowski, P. Parys, L. Ryc, J. Wolowski, E. Woryna, and A. B. Vankov, “Intensity-dependent characteristics of a picosecond laser-produced Cu plasma,” J. Phys. D: Appl. Phys. 34(12), 1885–1891 (2001). [CrossRef]
56. H. Nakano, A. A. Andreev, and J. Limpouch, “Femtosecond X-ray line emission from multilayer targets irradiated by short laser pulses,” Appl. Phys. B 79(4), 469–476 (2004). [CrossRef]
57. C. Reich, P. Gibbon, I. Uschmann, and E. Forster, “Yield optimization and time structure of femtosecond laser plasma K alpha sources,” Phys. Rev. Lett. 84(21), 4846–4849 (2000). [CrossRef]
58. D. Schick, M. Borchert, J. Braenzel, H. Stiel, J. Tümmler, D. E. Bürgler, A. Firsov, C. von Korff Schmising, B. Pfau, and S. Eisebitt, “Laser-driven resonant magnetic soft-x-ray scattering for probing ultrafast antiferromagnetic and structural dynamics,” Optica 8(9), 1237–1242 (2021). [CrossRef]
59. P. P. Rajeev and G. R. Kumar, “Polarization independent hot electron production from modulated solid surfaces,” Opt. Commun. 222(1-6), 9–15 (2003). [CrossRef]
60. P. P. Rajeev, S. Banerjee, A. S. Sandhu, R. C. Issac, L. C. Tribedi, and G. R. Kumar, “Role of surface roughness in hard-x-ray emission from femtosecond-laser-produced copper plasmas,” Phys. Rev. A 65(2002).
61. D. Metzner, P. Lickschat, and S. Weissmantel, “High-quality surface treatment using GHz burst mode with tunable ultrashort pulses,” Appl. Surf. Sci. 531, 147270 (2020). [CrossRef]
62. J. Uhlig, W. Fullagar, J. N. Ullom, W. B. Doriese, J. W. Fowler, D. S. Swetz, N. Gador, S. E. Canton, K. Kinnunen, I. J. Maasilta, C. D. Reintsema, D. A. Bennett, L. R. Vale, G. C. Hilton, K. D. Irwin, D. R. Schmidt, and V. Sundstrom, “Table-Top Ultrafast X-Ray Microcalorimeter Spectrometry for Molecular Structure,” Phys. Rev. Lett. 110(13), 138302 (2013). [CrossRef]
63. J. N. Ullom, J. A. Beall, W. B. Doriese, W. D. Duncan, L. Ferreira, G. C. Hilton, K. D. Irwin, C. D. Reintsema, and L. R. Vale, “Optimized transition-edge x-ray microcalorimeter with 2.4 eV energy resolution at 5.9 keV,” Appl. Phys. Lett. 87(19), 194103 (2005). [CrossRef]
64. A. Zymakova, M. Albrecht, R. Antipenkov, A. Spacek, S. Karatodorov, O. Hort, J. Andreasson, and J. Uhlig, “First experiments with a water-jet plasma X-ray source driven by the novel high-power-high-repetition rate L1 Allegra laser at ELI Beamlines,” J. Synchrot. Radiat. 28(6), 1778–1785 (2021). [CrossRef]
65. W. Fullagar, J. Uhlig, M. Walczak, S. Canton, and V. Sundstrom, “The use and characterization of a backilluminated charge-coupled device in investigations of pulsed x-ray and radiation sources,” Rev. Sci. Instrum. 79(10), 103302 (2008). [CrossRef]
