Laser pulses that have a high peak power are required in many research fields for studying nonlinear light-matter interactions, optical damage, plasma physics phenomena, etc. Stimulated Brillouin scattering (SBS) is widely used in laser systems to improve the peak power of laser pulses [1–3] and effectively compress nanosecond pulses down to sub-nanosecond laser pulses [4–6]. SBS features a significant small-signal gain [7], high signal-to-noise ratio [8], broad wavelength range [9], and the capability of pulse compression using a short-duration impulse amplification [10].

The pulse compression using an SBS has provided several achievements. For the application of Brillouin amplification to obtain a high-energy short pulse, the following factors should be considered. With the increase of the pump energy, an optical breakdown will occur at the focusing position in the generation cell. This phenomenon does not allow higher pump energies for the SBS pulse compression. In addition, in the SBS, a backward Stokes pulse is generated from the noise [11,12], which causes a time-domain instability of the Stokes pulse.

A stimulated Brillouin amplification is introduced to overcome the above challenges. By introducing a preset Stokes seed, the self-phase of the acoustic wave in the SBS can be locked [13]. This improves the stability of the SBS. In this Letter, we report an active frequency matching method for an SBS amplification to achieve a high-power laser pulse. In addition, the efficiency of the Stokes light extraction is experimentally investigated for different frequency shifts. For studying the stimulated Brillouin amplification, it is important to ensure that the frequency shift between the Stokes and pump light is equal to the acoustic field frequency in the medium. This should be satisfied, as the energy extraction decrease in the Brillouin amplification is significant, owing to the phase mismatch between the pump and Stokes seed pulses. The details are discussed below.

The experiment in this Letter employs an Nd: glass laser system that has been reported by Li et al. [14]. This laser system consists of five subsystems including a front-end, preamplifier, main amplifier, frequency converter, and parameter diagnostic unit. The frequency modulation is performed at the front-end system. We obtained a 200 ps Stokes pulse and 5 ns square pump pulse using an amplitude modulator (Mach–Zehnder modulator) that is controlled by an arbitrary waveform generator. These two pulses are separated into two fibers using a splitter. The 200 ps laser pulse is modulated using a down-frequency shift, achieved using an acouso-optic modulator. After the frequency modulation, the pump and Stokes pulses are coupled in a stacking pulse and are amplified in the following amplifier.

Once the frequency is doubled using the potassium dihydrogen phosphate crystal, the output laser pulse enters the main optical system. The beam size is narrowed (half of the initial value) using two lenses: a convex lens that has a focal length of 1500 mm and concave lens that has a focal length of 750 mm. The Brillouin amplification system is based on a noncollinear structure. Compared with the collinear structure, this noncollinear structure does not require complex optical systems, i.e., the use of polarizers and wave plates for beam separation and coupling is not required. Thus, it can simplify the optical path and reduce the damage risk of the amplification system. Meanwhile, a noncollinear scheme is more conducive to SBS amplification with larger beam size and a larger-energy laser. The crossing angle between the Stokes and pump pulses is 2°. One of the challenges in the noncollinear amplification is the decrease of the Brillouin gain coefficient, which can be overcome using a frequency shift modulation.

The SBS medium used in experiments in FC-40 is a kind of perfluorinated amines, which is composed of $({\mathrm{C}}_4{\mathrm{F}}_9{)}_3\mathrm{N}$ and $({\mathrm{C}}_4{\mathrm{F}}_9{)}_2{\mathrm{NCF}}_3$. It has a Brillouin gain coefficient of 1.8 cm/GW with a phonon lifetime of 0.2 ns (at 1053 nm). A theoretical analysis shows that the energy extraction has the highest gain at a frequency shift of 1.55 GHz. Therefore, the frequency shift should be modulated to this value at the front-end shown in Fig. 1.

 

Fig. 1. Frequency modulation at the front-end.

Download Full Size | PDF

The experimental setup is shown in Fig. 2. The laser wavelength in experiments is 527 nm($2\omega$) to reduce the risk of optical damage. Before entering into the medium cell, the laser energy, near field, and waveform are measured by sampling using a wedge plate. The cell used in this experiment has a length of 600 mm and a cross section of $100\mathrm{mm}\times 80\mathrm{mm}$. After passing through the cell, the laser will re-enter the cell after reflections from two mirrors, so that the 200 ps pulse can extract energy from the 5 ns pulse. The input and output laser pulses are measured by sampling using a 3° wedge plate. The wedge plate is a piece of uncoated quartz. The measured light is sampled form its Fresnel reflection on both the front and the back surfaces. The reflected light of the front and back surfaces propagates in different directions and gets separated, because there is a 3° angle of the wedge plate. Thus, the energy, the waveform, and the beam profile can be measured. The calibration of the sampling rate has been completed before experiments.

 

Fig. 2. Experimental setup of noncollinear Brillouin amplification using an injected frequency downshifted Stokes seed pulse.

Download Full Size | PDF

The results were obtained using a wedge plate. There are two wedge plates in the experimental path; each of them is used for sampling the energy and waveform of the laser pulse, before and after the SBS, respectively. Their sampling rates are calibrated before the experiment.

The waveform is measured using photodiodes (Ultrafast UPD-50-UP) and displayed on an oscilloscope (Tektronix DPO71604B). The energy is measured using an energy meter (Ophir PE100BF-DIF ROHS). The measured energy corresponds to the sum of the Stokes and pump pulses, i.e., to the stacking pulse. The Stokes and pump energies can be calculated using the integrated area that can be obtained using the oscilloscope.

According to the above theoretical analysis, the frequency shift will affect the magnification and energy extraction efficiency of the SBS amplification. In this Letter, we performed an SBS amplification experiment for different frequency shifts that ranged from 1.05 to 2.05 GHz. The delay time between the Stokes and pump pulses is 10.5 ns. The beam profile in the experiments is shown in Fig. 3. The experimental result shows that the output near field modulation is 1.56: 1.

 

Fig. 3. Beam profile in experiments.

Download Full Size | PDF

The experimental results verify the importance of the frequency matching for the SBS amplification. If there is a mismatch between the Stokes and pump pulses, the Brillouin gain coefficient rapidly decreases, which can significantly reduce the power amplification and energy extraction. Therefore, in order to ensure a large SBS amplification, the frequency matching has to be satisfied.

Typical experimental results are shown in Fig. 4; the red curve represents the waveform of the stacking pulse before the amplification, while the blue curve represents the output after the amplification. Figure 4(a) shows the experimental results at a pump intensity of $234\mathrm{MW}/{\mathrm{cm}}^2$ for a frequency shift of 1.55 GHz. The 200 ps Stokes output pulse is amplified to a $2082\mathrm{MW}/{\mathrm{cm}}^2$; the output energy is 2.4 J. By calculating the integral of the waveform trace, the output Stokes pulse and pump pulse energy can be obtained. This result shows that the pump pulse energy is extracted. It can be noticed that the amplified 200 ps laser pulse is not broadened; however, it is compressed. This can improve the Stokes intensity in the SBS amplification. Figure 4(b) shows the results for the frequency shift of 1.05 GHz. The experimental results show that the output power of the Stokes light reaches a value of $\mathrm{1,351}\mathrm{MW}/{\mathrm{cm}}^2$. The extraction efficiency decreases owing to the mismatch in frequency shift.

 

Fig. 4. Waveform after the SBS amplification at a pump intensity of $234\mathrm{MW}/{\mathrm{cm}}^2$ for frequency shifts of (a) 1.55 and (b) 1.05 GHz.

Download Full Size | PDF

In the experiments, SBS amplification with different frequencies shifts between Stokes, and the pump is carried at the same pump intensity; the results are shown in Fig. 5. Figure 5(a) shows a theoretical simulation of the gain variations caused by the frequency shift between the Stokes and pump pulses at a crossing angle of 2°. It is proved that in the noncollinear Brillouin amplification, the frequency mismatch will cause the Brillouin gain to drop rapidly. The maximum Brillouin gain can be obtained when the wave vector is strictly matched. As shown in Fig. 5(b), in the SBS amplification experiment with different frequency shifts, the amplified Stokes pulse has the highest energy extraction efficiency in the condition when the frequency shift strictly matches the wave vector. While there exists frequency detuning between Stokes and the pump pulse, the energy extraction efficiency decreases with the Brillouin gain coefficient. The experimental results are in good agreement with the theoretical analysis.

 

Fig. 5. (a) Theoretical simulation of the normalized gain coefficient as a function of the frequency shift in an FC-40. (b) Energy extraction efficiency as a function of the frequency shift in experiments.

Download Full Size | PDF

The experimental results and analysis show that the proposed SBS amplification method that is based on an active frequency modulation can be employed to achieve a high-power laser pulse of hundreds of picoseconds. The frequency mismatch between the Stokes and pump pulses directly affects the extraction efficiency that decreases owing to the phase mismatch.

A novel technique for the generation of high-energy laser pulses of hundreds of picoseconds by using an SBS in high-power solid-state laser systems was proposed. Frequency modulation between the Stokes and pump pulses was achieved at the front-end using acousto-optic devices. Through precise frequency matching, an energy extraction efficiency larger than 60% could be achieved. If the SBS medium or crossing angle is changed, the frequency shift can be modified to obtain a high-energy extraction efficiency.