With recent progress in chirped pulse amplification (CPA) [1] and titanium-doped sapphire (Ti:sapphire) media, extremely high laser peak powers of petawatt (PW) to multi-PW [2–4] and high focused intensities of ${{10}^{22}}\;{\rm W}/{{\rm cm}^2}$ or more [5–7] have been made available. Recently, a 10 PW Ti:sapphire laser has been developed in Romania under the European Extreme Light infrastructure (ELI) program [8]. The Station of Extreme Light (SEL) in China, which plans to output 100 PW, has also been designed and will be constructed in 2024 [9]. Focused intensities of ${{10}^{23}}\;{\rm W}/{{\rm cm}^2}$ or more will be achieved in the near future. Such lasers have proven to be a powerful drive source to provide opportunities for experimental investigation of the behavior of matter in ultra-high electromagnetic fields in small-scale laboratories.

Any pulses with intensities of ${{10}^{10}} - {{10}^{11}}\;{\rm W}/{{\rm cm}^2}$ prior to the main pulse are capable of ionizing solid targets, which will then significantly influence the laser-plasma interaction process [10]. Therefore, because an increasing number of facilities are providing high focused intensities, the temporal contrast has become one of the major issues in all high-intensity laser-plasma experiments over the last two decades. Regarding the temporal laser pulse shape, three different challenges are known: amplified spontaneous emission (ASE) [11], intensity spikes prior to the main peak called pre-pulses [12], and a slowly rising slope of tens of picoseconds around the main pulse. ASE and the pre-pulse may have major influences, even nanoseconds before the main pulse. The last one, which has a non-negligible effect on the plasma physics processes occurring on the target, dominates at short time scales and is caused by the surface roughness of the optics in the stretcher and compressor [13,14] and by the nonlinear transfer of characteristic post-pedestal to pedestal [15]. The slowly rising slope is difficult to remove by a laser itself, for example, without plasma mirrors [16].

Several technologies have been developed to improve the temporal contrast of high-intensity laser systems, in particular, double CPA (DCPA) [17], saturable absorption [18], cross-polarized wave generation (XPW) [19], and optical parametric amplification (OPA) [20]. A combination of DCPA, saturable absorption or XPW, and OPA allows a practical solution, in particular, for ASE, supporting an ASE contrast of ${{10}^{11}} - {{10}^{12}}$, even at PW power levels. By greatly enhancing the ASE contrast, pre-pulses that have been previously buried by ASE have become more visible recently.

In order to remove the pre-pulses, it is also important to eliminate post-pulses. The post-pulses do not influence laser matter interaction experiments in a direct way. However, when the stretched pulse duration in a CPA system is longer than the time difference between main and post-pulses, the two stretched pulses interfere with each other and lead to spectral interference. This sinusoidal modulation of the spectra is equivalent to a sinusoidal intensity modulation within the temporal shape of the stretched main pulse. Due to the intensity dependence of the B-integral accumulated by the main laser pulse in the subsequent optical elements of the laser chain, the intensity modulation can again cause a sinusoidal modulation in the spectral phase of the stretched main pulse. The modulated electric field nonlinearly changes the optical parameters of the optical elements. The time-dependent optical parameters, especially for the active medium, modulate the electric field more. This causes modulation in both the phase and the intensity of the field. After compression, a pre-pulse is therefore generated from a post-pulse [12,21,22]. These pre-pulses, which can have relatively high intensities, will play a key role in the experiments with high focused intensities. Therefore, the behavior of the pre-pulses, along with the accumulated B-integral from the main pulse, should be investigated, and the pre-pulses should be removed.

In this Letter, we report the experimental observation and investigation of the behavior of the pre-pulses generated by post-pulses in the laser chain of the PW-class J-KAREN-P laser facility. We have found that the generated pre-pulses are greatly delayed and distorted in time, and the peak intensities are drastically suppressed when the time delay between the stretched and post-pulses in the CPA system is large. We also briefly describe the identification of the origins of the pre-pulses by post-pulses and removal of the pre-pulses in the J-KAREN-P experimentally. This is the first observation and investigation of such pre-pulses using a PW laser facility, to the best of our knowledge.

For these tests, we have used the J-KAREN-P facility at the Kansai Photon Science Institute, National Institutes for Quantum and Radiological Science and Technology, Japan. A detailed description of the J-KAREN-P can be found elsewhere [18]. The J-KAREN-P is based on an OPCPA/Ti:sapphire hybrid DCPA architecture linked by a saturable absorber. The output pules with 25 fs pulse duration from the first CPA stage are passed through a saturable absorber to initially enhance the temporal contrast. The laser pulses are temporally expanded by the stretcher. Downstream of the stretcher, an acousto-optic programmable dispersive filter (AOPDF) controls the spectral phase prior to further amplification, and a small-aperture Faraday isolator blocks the backward pulses from the downstream system. The stretched pulses are first amplified and broadened spectrally in the OPCPA pre-amplifier, pumped by a frequency-doubled Nd:YAG laser (Amplitude Laser Group Continuum, Intrepid). The pump laser delivers an arbitrary temporal pulse shape with a programmable optical pulse shaper consisting of the Mach–Zehnder modulator, bias control circuit, and a programmable arbitrary pulse synthesizer. The pulses are passed through a fast Pockels cell to remove the nanosecond-order ASE pedestal; then they are amplified in subsequent Ti:sapphire amplifiers, namely a pre-amplifier, a cryogenically cooled power amplifier, and two booster amplifiers. Another large-aperture Faraday isolator is also employed between the Ti:sapphire pre- and power amplifiers to prevent back reflection pulses from the downstream system. The pulses are finally recompressed in a compressor. The pulse duration is optimized by minimizing the spectral phase distortion through a feedback loop between the AOPDF and a real-time spectral phase measurement device (Fastlite, Wizzler). The stretched pulse duration and the recompressed pulse duration are around 0.5 ns and 40 fs, respectively, in this experiment.

Two plane-parallel plates made of uncoated fused silica with different thicknesses of ${{\sim}}{1}$ and ${\sim}{11.8}\;{\rm mm}$ were inserted into the J-KAREN-P laser chain, resulting in a reflection of ${\sim}{3.4}\% $ per surface to generate post-pulses by a double reflection of the transmitted pulse at the two surfaces. The plane-parallel plates are inserted just before the Ti:sapphire pre-amplifier; thus, the Ti:sapphire amplifiers described above are seeded with a main pulse accompanied by these created post-pulses. Before the experiment, in the absence of nonlinear effects, when the plane-parallel plates were placed right in front of the third-order cross-correlator (Amplitude Technology, Sequoia) for the calibration confirmation, the contrast of the post-pulses with these plane-parallel plates was close to the level due to Fresnel reflection, ${\sim}{{10}^{ - 3}}$. The third-order cross-correlator was usable for both 10 and 0.1 Hz by just changing the trigger signal. The artificial pre-pulses appear in the third-order cross-correlation measurement due to the mixing of the second harmonics of the post-pulses and the fundamental of the main pulse. The contrast of these measurement artifacts was confirmed to be equal to ${\sim}{{10}^{ - 6}}$, the square of the post-pulse contrast. These plane-parallel plates therefore lead to post-pulses with the contrast of ${\sim}{{10}^{ - 3}}$ at a time $t$ determined by $t={2}dn/c$, where the $d$ is the plate thickness, $n$ is the refractive index of the plane-parallel plate, and $c$ is the speed of light. Typical values are $d\,=\,\sim{1}$ and ${\sim}{11.8}\;{\rm mm}$ which, respectively, leads to $t\,=\,\sim{10}$ and ${\sim}{118}\;{\rm ps}$ for $n\,=\,\sim{1.45}$.

 

Fig. 1. Measured contrast of (a) post-pulses and (b) pre-pulses in the J-KAREN-P seeded with a pulse accompanied by a post-pulse reflected off the plane-parallel plate with ${\sim}{1}\;{\rm mm}$ thickness: the calibration (plane-parallel plate is placed after the compressor) (green line), output energies of ${\sim}{45}\;{\rm mJ}$ (blue line), ${\sim}{1.8}\;{\rm J}$ (red line), and ${\sim}{26}\;{\rm J}$ (black filled circles). These energies correspond to the B-integrals of ${\sim}{0.037}$, ${\sim}{0.25}$, and ${\sim}{0.85}\;{\rm rad}$, respectively.

Download Full Size | PDF

Figure 1 shows the measured contrast of the J-KAREN-P seeded with a pulse accompanied by a post-pulse reflected off the plane-parallel plate with ${\sim}{1}\;{\rm mm}$ thickness. In this case, ${\sim}{98}\% $ of the stretched post-pulse overlaps with the stretched main pulse temporally. The post-pulses at ${\sim}{10}\;{\rm ps}$ of ${\sim}{{10}^{ - 3}}$ contrast have been generated due to the Fresnel reflection. The generation of the pre-pulse ${\sim}{10}\;{\rm ps}$ before the main pulse is also visible. The pre-pulses are generated from a post-pulse at an identical time difference before the main pulse. The green line shows the calibration which means that plane-parallel plate is placed after the compressor (right in front of the third-order cross-correlator). The pre-pulse temporal contrasts are degraded by increasing the pulse energy, while keeping the post-pulse contrasts. The different peak pre-pulse contrasts of ${\sim}{4.1} \times {{10}^{ - 6}}$ (blue line), ${\sim}{5.3} \times {{10}^{ - 5}}$ (red line), and ${\sim}{6.8} \times {{10}^{ - 4}}$ (black filled circles) are due to three different energies at the exit of the operating amplifiers of ${\sim}{45}\;{\rm mJ}$, ${\sim}{1.8}\;{\rm J}$, and ${\sim}{26}\;{\rm J}$, respectively. These energies correspond to the B-integral values of ${\sim}{0.037}$, ${\sim}{0.25}$, and ${\sim}{0.85}\;{\rm rad}$, respectively, in this experiment. The B-integral values have been estimated taking the optics in the J-KAREN-P laser chain into account. When the B-integral values are higher, the pre-pulse intensities are higher. The peak intensity contrasts of the pre-pulses generated by the post-pulse can be estimated [12] to be ${4.6} \times {{10}^{ - 7}}$, ${2.0} \times {{10}^{ - 5}}$, and ${2.4} \times {{10}^{ - 4}}$. The theoretical prediction fit the experimental data reasonably well within a factor of ${2.6 - 8.9}$.

Figure 2 shows the measured contrast of the J-KAREN-P seeded with a pulse accompanied by a post-pulse reflected off the plane-parallel plate with ${\sim}{11.8}\;{\rm mm}$ thickness. In this case, ${\sim}{75}\% $ of the stretched post-pulse overlaps with the stretched main pulse temporally. The post-pulses at ${\sim}{118}\;{\rm ps}$ with ${\sim}{{10}^{ - 3}}$ contrast have been generated due to the Fresnel reflection. However, a completely different behavior of the pre-pulses from the thin ${\sim}{1}\;{\rm mm}$ plane-parallel plate, which has a small time delay between the main and post-pulses, has been observed. The generation of the pre-pulse ${\sim}{118}\;{\rm ps}$ before the main pulse is also visible and is an artifact. The green line also shows the calibration which means that a plane-parallel plate is placed after the compressor. New real pre-pulses by the post-pulse have appeared, and the peak pre-pulse contrasts are not significantly degraded, even if increasing the pulse energy, while keeping the post-pulse contrasts. An interesting feature is that the pre-pulses are not generated from a post-pulse at an identical time difference before the main pulse. Namely, the peak of the pre-pulses generated by post-pulse is delayed by ${\sim}{4.8}\;{\rm ps}$, and the temporal shape is greatly distorted to an asymmetric broadened one. Another interesting feature is that the peak pre-pulse contrast is drastically suppressed. Namely, the smaller different peak pre-pulse contrasts of ${\sim}{1.8} \times {{10}^{ - 8}}$ (blue line), ${\sim}{2.0} \times {{10}^{ - 6}}$ (red line), and ${\sim}{3.8} \times {{10}^{ - 6}}$ (black filled circles) are due to three different energies at the exit of the operating amplifiers of ${\sim}{45}\;{\rm mJ}$, ${\sim}{1.8}\;{\rm J}$, and ${\sim}{26}\;{\rm J}$, respectively. These energies correspond to the B-integral values of ${\sim}{0.037}$, ${\sim}{0.25}$, and ${\sim}{0.85}\;{\rm rad}$, respectively, where the B-integral values have been also estimated taking the optics in the J-KAREN-P laser chain into account. When the B-integral values are higher, the intensities and pulse durations of the delayed pre-pulses are higher and broader, respectively. The experimental measurements of the pre-pulse peak intensity are much smaller than the theoretical prediction by a factor of 0.039–0.0158.

 

Fig. 2. Measured contrast of (a) post-pulses and (b) pre-pulses in the J-KAREN-P seeded with a pulse accompanied by a post-pulse reflected off the plane-parallel plate with ${\sim}{11.8}\;{\rm mm}$ thickness: the calibration (plane-parallel plate is placed after the compressor) (green line), output energies of ${\sim}{45}\;{\rm mJ}$ (blue line), ${\sim}{1.8}\;{\rm J}$ (red line), and ${\sim}{26}\;{\rm J}$ (black filled circles). These energies correspond to the B-integrals of ${\sim}{0.037}$, ${\sim}{0.25}$, and ${\sim}{0.85}\;{\rm rad}$, respectively.

Download Full Size | PDF

The pre-pulses generated by the post-pulse from the thin ${\sim}{1}\;{\rm mm}$ plane-parallel plate can be explained by the accumulation of the nonlinear phase [12]. However, the behavior of greatly delayed and asymmetrically broadened pre-pulses by the post-pulses from the thick ${\sim}{11.8}\;{\rm mm}$ plane-parallel plate is interesting and not understood yet. In this case, the time delay of ${\sim}{118}\;{\rm ps}$ between the main and post-pulses is significantly larger, which means that only ${\sim}{75}\% $ of the main and post-pulses interfere with each other in our system. The smaller spectral interference part generates a pre-pulse with broader pulse duration because of its narrower spectral bandwidth. The dispersion of the pre-pulses is far different from that of the main pulse. In particular, the different positive second- and third-order dispersion of the pre-pulse that is accumulated in the laser system is not allowed to recompress simultaneously with the same setup of the stretcher-compressor configuration which is optimized for the main pulse. Consequently, the generated pre-pulses have much broader pulse duration and a significantly longer foot starting at around the peaks of the pre-pulses with a decay due to the positive odd-order dispersion, also likely in combination with self-phase modulation. Due to the narrower spectral bandwidth and the uncompensated large dispersion with the self-phase modulation, the generated pre-pulses are asymmetrically broadened and distorted. As a result, the peak intensity of the pre-pulse is lower than that compared with the thin ${\sim}{1}\;{\rm mm}$ plane-parallel plate case. In addition, the spectral intensity modulation part of the main pulse is in the blue region in wavelength, which is the later region of the main pulse in time in the positively chirped laser system. The phase in only the blue region is asymmetrically modulated. Therefore, the generated pre-pulses are delayed in time. Further theoretical investigation of this nonlinear coupling effect is needed to fully understand and evaluate the degradation mechanism and will be presented in future work.

Figure 3(a) shows the measured contrast of the J-KAREN-P laser system with ${\sim}{1}$ and ${\sim}{10}\;{\rm J}$ output pulse energies. According to our above experience and knowledge, we have identified real and artificial pre-pulses generated by the post-pulses.

 

Fig. 3. (a) Previous contrast of the J-KAREN-P laser system with ${\sim}{1}$ (solid red line) and ${\sim}{10}\;{\rm J}$ (black filled circles) output pulse energies. (b) Current contrast of the J-KAREN-P laser system by employing optical components with a small wedge angle with ${\sim}{1}$ (solid red line) and ${\sim}{10}\;{\rm J}$ (black filled circles) output pulse energies.

Download Full Size | PDF

The real pre-pulses are coming from the post-pulses due to double reflection in the Ti:sapphire crystal of the power amplifier, the Ti:sapphire crystal of the pre-amplifier, the small-aperture Faraday isolator, the windows of the cryogenically cooled power amplifier, and the optics inside the oscillator.

In order to remove the post-pulses from the Ti:sapphire crystals and windows, we introduced new crystals and windows with a small wedge of 0.5 deg. The wedge was in the horizontal plane, so that the angular dispersion and pulse front delay were cancelled completely in four-pass geometry of the amplifier with two steering mirrors after each path. The effect of the wedged optics on focal spot was also estimated using Zemax software, considering the angular dispersion and was found to be nearly negligible. Actually, the same focusability was confirmed experimentally. The angular dispersion in the three-pass or five-pass geometries can also be removed with the addition of an extra wedge plate which compensates for the residual dispersion in the geometries. As the current laser system has sufficient isolation from back reflection light, the small-aperture Faraday isolator was removed from the laser chain to eliminate the post-pulses. Since the pre-pulse at $ - {40}\;{\rm ps}$ is generated by the post-pulse from the commercial oscillator, it is difficult to avoid this pre-pulse. Therefore, as a future work, the main pulse from the oscillator will be amplified selectively by a few picosecond short pulse laser pumped OPCPA, a specially designed home-made oscillator will be fabricated, and plasma mirrors will be installed after compression, to suppress and remove this pre-pulse. By introducing the wedged optics, the temporal contrast is measured with ${\sim}{1}$ and ${\sim}{10}\;{\rm J}$ output energies to investigate the pre-pulses by the post-pulses. The pre-pulses except a pre-pulse at $ - {40}\;{\rm ps}$ have been removed, as shown in Fig. 3(b).

In conclusion, we have experimentally investigated pre-pulse generation by post-pulses due to the nonlinear coupling associated with the B-integral at the PW facility J-KAREN-P. In the case of a small time delay of ${\sim}{10}\;{\rm ps}$ (${\sim}{98}\% $ overlap) between the stretched main and post-pulses, when introducing a thin ${\sim}{1}\;{\rm mm}$ plane-parallel plate, the pre-pulses generated by the post-pulse appeared at the same time as the artificial pre-pulse of their counterpart post-pulse in time. In contrast, it has been found that in the case of a large time delay of ${\sim}{118}\;{\rm ps}$ (only ${\sim}{75}\% $ overlap) between the stretched main and post-pulses, when introducing a thick ${\sim}{11.8}\;{\rm mm}$ plane-parallel plate, the pre-pulses generated by the post-pulse are greatly delayed by ${\sim}{4.8}\;{\rm ps}$ from the artificial pre-pulse of their counterpart post-pulse in time and are significantly distorted to an asymmetrically broadened one. The peak intensities of the pre-pulses are drastically suppressed due to the distortion in time. According to our knowledge and experience based on this investigation, we have identified the origins of the pre-pulses generated through nonlinear coupling by the post-pulse. By introducing small wedged optics, we have demonstrated elimination of the pre-pulses at our J-KAREN-P PW facility. Since a 10 PW laser was demonstrated more recently, and additional 10 PW lasers, even a 100 PW laser, are under construction, the focused intensities are rapidly increasing. We believe that the results of this investigation described here, therefore, should be useful guidelines for the characterization and improvement of pre-pulse contrast and be of great significance for ultra-high peak power and ultra-high-intensity CPA-based lasers.