2.1.

Pump Laser at 1064 nm

The pump laser is an Nd:YAG-MOPA laser,^7,8 which can be injection-seeded and cavity-controlled to achieve single-longitudinal mode (SLM) operation. The spectral, temporal, and the main spatial beam properties are generated in the low-energy Q-switched Nd:YAG-based oscillator. The pulse energy of 7.5 mJ from the oscillator is amplified in two subsequent Innoslab-based amplifier stages. The system delivers up to 500 mJ of output pulse energy at $1\mu \mathrm{m}$, a repetition rate of 100 Hz and a pulse duration of 30 ns. The output beam is shaped using a cylindrical telescope to obtain a symmetrical beam with a beam quality $M^2$ of $1.8\times 1.4$.

2.2.

Frequency Conversion Unit

To avoid back reflections into the amplifier, a Faraday isolator is inserted into the beam path. About 20 mJ of the pump beam is extracted to pump the OPO. The beam is shaped with a spherical telescope to achieve a radius that guarantees an efficient working point of the OPO. The OPO setup is inherited from the MERLIN setup.¹¹ In the four-mirror ring cavity, potassium titanyl phosphate (KTP) was used as the nonlinear medium. All four mirrors had a high transmission of the idler wave to reduce possible thermal effects and back conversion at high intensities. The OPO was injection-seeded by a distributed feedback diode laser at 1645 nm through the signal outcoupling mirror. The cavity length of the OPO can be varied by a piezoelectric element and is actively controlled by a heterodyne detection scheme to guarantee SLM operation. The procedure of this control method is explained in Ref. 9.

The output signal beam is shaped with a spherical telescope to match the pump beam size in the OPA. The OPO signal beam and the main part of the pump beam are then combined using a dichroic mirror, so that a good temporal overlap can be achieved between pump and signal beam in the OPA. A delay line is inserted into the beam path of the pump. The optical path length of the delay line is set in a way that the pulse build-up time of the OPO can be compensated.

To investigate the amplification properties of the OPA for different setups, the OPA consists of up to four 15-mm long KTP crystals, with the possibility of removing the idler beam between them. According to the work of Arisholm et al.,¹² this can improve beam quality and output power. The KTP crystals are cut in a critical phase-matching configuration ($\theta =76\mathrm{deg}$, $\varphi =0\mathrm{deg}$). After the last OPA stage, the signal beam is separated from the residual pump and idler beam by dichroic mirrors.

3.1.

Optical Parametric Oscillator

The energetic output characteristic curves of the OPO are shown in Fig. 3. When it was injection-seeded and had an active cavity-length stabilization, the system achieved a maximum output energy of 8 from 21 mJ of the pump. This corresponds to an optical/optical efficiency of almost 40% and a quantum efficiency of almost 60%. The output bandwidth of the signal beam was measured using a heterodyne detection scheme and was around 40 MHz. Thus, the performance of the OPO is similar to that of the MERLIN setup.

Fig. 3

Output pulse energies and efficiency of the OPO stage over pump pulse energy.

The output pulse energy does not depend on the mode of operation of the pump laser, which can be operated in single-longitudinal frequency mode (seeded and with cavity-length stabilization) or in free-running mode (unseeded, no cavity-length stabilization).

The pulse duration of the OPO output was measured to be around 23 ns (full width at half maximum), which is 25% shorter than the pulse duration of the pump laser due to the build-up time of the OPO. The temporal shapes of the pump and signal pulse are shown in Fig. 4.

Fig. 4

Temporal shape of the pump pulses (black) and the OPO output pulses (red).

3.2.

Optical Parametric Amplifier

The amplification of the OPO signal output was measured for a different number of OPA stages. Figure 5 shows the OPA output pulse energies for one to four OPA stages. A maximum signal pulse energy of 111 mJ was achieved at 400 mJ of the pump energy. This corresponds to more than 26% conversion efficiency.

Fig. 5

Output pulse energies after a different number of OPA conversion stages. Inset: output beam profile at 1645 nm.

These results were obtained when the pump laser was operated without injection seeding and without cavity-length stabilization, hence running on multiple longitudinal modes. When the laser was operated with injection seeding and cavity-length stabilization, and hence running in SLM, the output energies of the OPA stages decreased as can be seen in Fig. 6 (here shown for three OPA stages).

Fig. 6

Comparison of output pulse energies for unseeded, free-running and seeded, longitudinally single-mode pump laser with three OPA stages.

Since the other input characteristics of the OPA, such as input signal pulse energy (see Fig. 3), beam profile, pulse width, and temporal delay between signal and pump, etc. were unchanged, the spectral input characteristics of the OPA were the only parameters that changed between the two different operating modes. This effect is well reproducible. Further investigations showed that the gain in every single OPA stage decreased by a factor of 1.5, if the pump laser is operated in SLM. Therefore, optimizing the output energy in SLM operation should be achievable due to adaptions of the beam radius and crystal length. This and further investigations on the reasons for this effect are currently being conducted.