Femtosecond petawatt-level laser pulses make it possible to experimentally investigate highly nonlinear processes in atomic, molecular, plasma, and solid-state physics, and to access previously unexplored states of matter [1,2]. The technique of chirped pulse amplification (CPA) has opened a new avenue to the production of ultrahigh energy and ultrashort duration pulses without optical damage to amplifiers and optical components [3]. Since its invention in the early 1980s, many scientists have been seeking to boost the laser power to several hundreds of terawatts. To date, several laboratories worldwide have held petawatt-level laser systems based on Ti:sapphire crystal and CPA technology [4–7]. However, further increase in amplified energy is limited in principle because of the parasitic lasing (PL) of large aperture Ti:sapphire [8]. Meanwhile, a new technique of optical parametric–chirped pulse amplification (OPCPA) was proposed [9], which has been regarded as a powerful scheme for obtaining high-power laser pulse in a short length crystal, a lower B-integral, and broadband gain bandwidth without significant spectral phase errors [10,11]. At the same time, OPCPA offers the possibility of implementing ultrashort pulse laser with wide frequency tunability. The capability of delivering multiterawatt power from a staged OPCPA system has been shown in small-scale OPCPA systems with standard-sized nonlinear crystals [12–15]. Petawatt-level OPCPA laser systems based on large-aperture ${\mathrm{KH}}_2\mathrm{PO}4$ (KDP) or ${\mathrm{KD}}_2\mathrm{PO}4$ (DKDP) nonlinear crystals were also obtained. Presently, OPCPA systems have generated energies as high as 35 J near 1053 nm based on a DKDP crystal with a potential power of 300 TW [16], and 38 J near 910 nm based on a DKDP crystal with a peak power of 0.56 PW [17]. Additionally, the combination of OPCPA and CPA is already used [18], in which the OPCPA is used as preamplifier and the Ti:sapphire CPA as boost amplifier, but the PL presented here is unavoidable in this configuration.

Although the PL in large Ti:sapphire amplifiers limits the higher energy amplification of CPA laser systems, the moderate-scale Ti:sapphire CPA laser can deliver a broadband and stable laser pulse with high conversion efficiency at the output energy of several tens of joules. However, a large-scale OPCPA can support a higher energy amplification without PL and significant spectral errors. As we know, ${\mathrm{LiB}}_3{\mathrm{O}}_5(\mathrm{LBO})$ is an attractive nonlinear crystal that can support high efficiency and broadband OPCPA near 800 nm. Its recent development for larger size growth makes it possible for high-energy amplification [19]. All these factors indicate that a high peak power laser system can be produced by combining a Ti:sapphire CPA front end and a LBO-OPCPA booster amplifier. The quite mature and stable Ti:sapphire CPA front end, operating near 800 nm, greatly reduces the complexity of an all-stage OPCPA petawatt laser system centered at other longer wavelength. Meanwhile, the narrower pulse width could be compressed with same spectral bandwidth. Additionally, OPCPA booster amplifiers based on nonlinear crystal can avoid the negative influence of gain narrowing of an all-stage CPA laser. Therefore, a hybrid system with a CPA front end and an OPCPA final amplifier is a potential design for building compact 10 PW laser.

For our design of a hybrid CPA-OPCPA laser system, the amplification based on nonlinear crystal in a noncollinear geometry can match the wavelength range of the Ti:sapphire CPA front end. In this Letter, we experimentally show a hybrid Ti:sapphire-CPA and LBO-OPCPA laser system, which produce an amplified energy of 28.68 J with spectral bandwidth of 80 nm (FWHM), what the authors believe to be the highest energy amplified in noncollinear LBO-OPCPA centered at 800 nm. Finally, the system demonstrates the compression to 0.61 PW with 33.8 fs pulse duration.

The basic design of this system (Fig. 1) relies on an idea of combination of CPA and OPCPA technologies. A commercial 75 MHz, sub-20 fs Ti:sapphire oscillator [not carrier-envelope phase (CEP) stable] was used to provide seed pulses for an aberration-free all-reflective Öffner-triplet-type stretcher with a $1200\text{groove}/\mathrm{mm}$ grating. After stretching, the laser pulse was sequentially amplified in the regenerative amplifier and two Ti:sapphire multipass amplifiers such that its energy reached $\sim 3\mathrm{J}$ with temporal duration of 1.6 ns (FWHM) and spectrum range from 758 to 840 nm at 5 Hz. Then the amplified laser pulse was up-collimated to a 54 mm diameter for injection. Finally, the injected signal energy of the OPCPA booster amplifier was stabilized at 2.89 J by removing the transmission and reflection losses of the optical components.

 

Fig. 1. Schematics of the hybrid CPA-OPCPA laser system.

Download Full Size | PDF

The OPCPA booster amplifier was pumped by the second-harmonic generation of an Nd:glass laser. The pump laser, shown in the largest shaded box, was seeded by a laser pulse sliced out of a continuous wave single longitudinal mode (CW-SLM) laser at 1 Hz. This pulse was amplified from $\sim 280\mathrm{pJ}$ to 3 mJ by a diode-pumped Nd:YLF regenerative amplifier. Then, the output pulse was temporal shaped by a high-speed electro-optic modulator with an adjustable rising edge to compensate for the pulse-shape transformation caused by the gain saturation in the subsequent Nd:glass amplifiers. Meanwhile, it was spatially apodized to produce a nearly flat spatial profile. Subsequently, the laser pulse was amplified in the following Nd:glass amplifiers. After frequency doubling, the second harmonic was down-collimated to 55 mm diameter. Finally, pump energy of up to 102 J with nearly flat spatial-temporal profiles at 526.5 nm was available in 2.89 ns pulse duration at a repletion rate of one shot every 20 min.

The noncollinear phase-matching scheme was used to obtain a broadband and energetic parametric amplification. The noncollinear geometry of the OPCPA stage leads to a spatial separation of signal, pump, and idler, and no further optics are needed to separate the beams as in collinear geometries or CPA’s. Presently, the nonlinear crystals that can be grown into larger sizes are LBO, yttrium calcium oxyborate (YCOB), and KDP (or DKDP) [20]. The main characters for noncollinear OPCPA at the central wavelength of 800 nm with pump of 526.5 nm are modulated, as shown in Table 1. Evidently, the YCOB crystal has the maximum value of the effective nonlinear coefficient ($d_{\text{eff}}$). Meanwhile, the gain bandwidths ($\mathrm{\Delta }\lambda$) of LBO and YCOB are much larger than that of DKDP (96% deuterium replacement for hydrogen). Additionally, the LBO crystal supports a better overlap of pump and signal beams than YCOB, arranged in a walk-off compensating scheme, which guarantees the higher conversion efficiency to the signal. Furthermore, the gain spectrum based on LBO crystal is not very sensitive to the phase-matching angles and noncollinear angles.

Table 1. Nonlinear Optical Parameters (${\lambda }_{\text{pump}}=526.5\mathrm{nm}$, ${\lambda }_{\text{signal}}=800\mathrm{nm}$)^a

View Table

LBO crystal was grown by the Top-Seeded Solution Growth method in the Technical Institute of Physics and Chemistry, CAS [19]. An inclusion free LBO crystal has been successfully grown and cut into an optical element with size of $80\mathrm{mm}\times 80\mathrm{mm}\times 12\mathrm{mm}$ for critically phase-matched type I ($\theta =90°$, $\phi =13.85°$). Both faces were polished and then coated for antireflection at pump and signal wavelengths.

The signal and pump pulses were electronically synchronized using a master clock circuit. Amplification was arranged in noncollinear geometry with 1.26° noncollinear angle between the pump and signal beams in the crystal. The beam diameters of the pump and the signal were 55 and 54 mm, respectively, which maintained a decent overlap in LBO. The injected spectrum of signal was shaped with a hole (Fig. 2) in the center by adjusting the tilt angle of the spectral shaping filter (Alpine Research Optics), inserted into the regenerative amplifier, to avoid the energy transferring from signal and idler to pump in high pump density (the detailed analysis will be presented in our following papers). Efficient OPCPA operation with a broadband gain and high conversion efficiency was achieved at optimized settings of the phase-matching angle, noncollinear angle and the location of the crystal for the overlap area of the pump and the signal. At the optimized conditions and injected signal energy of 2.89 J, the maximum amplified output was up to 28.68 J with pump energy of 101.6 J, corresponding to conversion efficiency of 25.38%, whereas the pump intensity was $\sim 1.49\mathrm{GW}/{\mathrm{cm}}^2$, as shown in Fig. 3. The conversion efficiency as a function of pump energy tends to reach the maximum at this pump intensity, and the amplified energy could be improved with the increase of pump energy. In addition, the corresponding spectrum still presented an evident hole, which also confirms that no backconversion occurred in the parametric amplification. Meanwhile, the red part was amplified with higher gain, which means that the FWHM spectral width was 80 nm by maintaining the spectrum full width. Due to the favorable phase-matching conditions, the FWHM of the amplified spectrum was broader than that of seed. With the gain attributed to the booster OPCPA amplifier of less than 10 and the high injected seed density of $\sim 0.13\mathrm{J}/{\mathrm{cm}}^2$, the side parametric fluorescence was not observed in the amplification process.

 

Fig. 2. Amplified spectrum after the booster amplifier (solid curve) and the seed signal spectrum (dashed curve).

Download Full Size | PDF

 

Fig. 3. Conversion efficiency as a function of pump power density ($\mathrm{Gw}/{\mathrm{cm}}^2$) and the output amplified signal energy dependence of the pump energy (J) with represented by solid curve and dashed curve, respectively.

Download Full Size | PDF

To test the stability of the hybrid CPA-OPCPA laser system, the amplified signal energy of the OPCPA booster amplifier was continuously recorded with the pump and signal energies fixed to 96 and 2.89 J, respectively. In thirteen single-shot measurements, the maximum energy was 27.14 J and the minimum energy was 25.18 J, corresponding to a fluctuation of less than 2.1% (RMS), as shown in Fig. 4(b). The measured spatial beam profile of the amplified signal is shown in Fig. 4(a), which indicates that a similar distribution of injection was maintained.

 

Fig. 4. (a) Measured spatial beam profiles of the amplified seed at the energy level of 28.68 J. (b) Measured output energy for the 96 J pump energy and 2.89 J injected seed energy. For all the measured single-shot, the maximum and minimum energy were 27.14 and 25.18 J, respectively. The RMS was about 2.1%.

Download Full Size | PDF

The amplified pulse from the OPCPA booster amplifier was recompressed by a grating compressor consisting of four $1480\text{groove}/\mathrm{mm}$ gold-coated holographic gratings after being expanded to 150 mm in diameter through an achromatic beam expander. After compression, the measured single-shot autocorrelation trace was clean, as shown in Fig. 5. The pulse duration was 33.8 fs corresponding to the Fourier transform limited duration of 27.3 fs, which is supported by the amplified spectrum in Fig. 2. Due to the absence of active dispersion compensation devices, the compressor did not precisely compensate the high-order dispersion. Thus, the single-shot autocorrelation was measured with small pre or postpulses. The compressor throughput efficiency was measured to be over 72%, such that the final output energy was 20.65 J, corresponding to peak power of 0.61 PW. Moreover, the clear aperture of the LBO crystal was larger than the beam size of the pump and the signal, which produced a potential intensity of more than one petawatt by scaling up the beam size of the pump and the signal with the same power density.

 

Fig. 5. Measured single-shot autocorrelation traces of a compressed amplified laser pulse.

Download Full Size | PDF

In conclusion, a hybrid CPA and OPCPA system with broadband amplified energy of 28.68 J and 25.8% conversion efficiency was demonstrated. The amplified spectrum range cover the full width of the injected signal from 758 to 840 nm. The improvement of output energy and the extension of the spectrum can be achieved by higher pump energy and wider injected spectrum, respectively. To the best of our knowledge, the result we obtained is the highest energy achieved so far with the OPCPA technique based on LBO crystal at the center wavelength of 800 nm. Although the maximum energy of the presented OPCPA system is still well below the state-of-the-art pure Ti:sapphire CPA system described in [7], the framework of a hybrid CPA-OPCPA laser system provides a new approach to achieve several tens of petawatts laser system, based on the development of a much larger optical aperture of nonlinear crystal and the available several tens of joules Ti:sapphire CPA front end.