**3. Results and Discussion**

The input-output characteristics in Q-switched operation are shown in Figure 2. The maximum output energies of 22.0 mJ/pulse, 19.1 mJ/pulse, 17.7 mJ/pulse and 15.1 mJ/pulse were obtained with no saturation when the repetition rates were 100 Hz, 150 Hz, 200 Hz and 250 Hz, respectively. According to the theoretical analysis of transient thermal distribution of repetitively pumped laser by W. Koechner, the temperature in the center of the laser rod increases with the increase of the repetition rate [24]. Serious heat accumulation at higher repetition rate will broaden the fluorescent spectrum and shorten the upper-level lifetime, leading to the decrease of laser output. The higher pump current was not tested to avoid the damage to the Nd:YAG crystal. The lasing thresholds were about 45 A for different repetition rates. The spatial profile of pump beam at the front surface of BGSe crystal was measured as shown in the inset of Figure 2 (linear in spatial dimensions and color scale). The 1064 nm pump beam diameter was 2.5 mm (vertical direction) × 1.0 mm (horizontal direction). The spot size in horizontal direction was shorter than in vertical direction due to the aberration caused by the Brewster polarizer.

The input-output characteristics of the generated mid-infrared wave at 4.06 μm with different repetition rates are shown in Figure 3a. With the increase of repetition rate, the LIDT decreased rapidly. When the repetition rate was above 150 Hz, damage on the surface of the BGSe crystal was observed with pump energy higher than 9 mJ/pulse (corresponding to peak intensity of about 20 MW/cm2). For higher repetition rate of 500–1000 Hz, optical damages were much more intense, which limited the further improvement in the repetition rate of the BGSe-OPO. The LIDT is lower than our recent report on the low repetition rate (10 Hz) experiment [25]. The thermal effect caused by the extra crystal absorption to the pump wave and the relatively low beam quality from the side-pumped laser were other possible reasons.

**Figure 2.** Input-output characteristics of side-pumped Nd:YAG laser. Inset: spatial profile of pump beam at the front surface of BGSe crystal.

**Figure 3.** (**a**) The input-output characteristics of the BGSe-based DP-SRO with different repetition rate; (**b**) The conversion efficiency of the BGSe-based DP-SRO with different repetition rate.

The maximum output energy of 1.28 mJ/pulse was achieved at pump energy of 10 mJ/pulse with a repetition rate of 100 Hz. The maximum output average power was 250 mW with a repetition rate of 250 Hz. A slight increase in the OPO threshold was observed with the increase of the repetition rate. The OPO thresholds were measured to be 2.65 mJ/pulse, 3.34 mJ/pulse, 3.54 mJ/pulse and 3.61 mJ/pulse for the repetition rate of 100 Hz, 150 Hz, 200 Hz and 250 Hz, respectively.

Figure 3b shows the conversion efficiencies of the BGSe-OPO with different repetition rates. The maximum conversion efficiency of 12.8% was obtained for the repetition rate of 100 Hz. For different repetition rates of 100 Hz, 150 Hz, 200 Hz and 250 Hz, the slope efficiencies were measured to be 16.4%, 17.0%, 14.4% and 16.6%, respectively. No saturation phenomenon was observed because the low pump peak intensity was limited by the LIDT. Further optimization to pump beam quality could improve the output characteristics of the BGSe-OPO.

The output energy and conversion efficiency of BGSe-OPO decreased with the increase of repetition rate due to the severe thermal effect. It is especially significant for the midinfrared crystal with poor thermal conductivity like AgGaS2 (1.4 W m−<sup>1</sup> K−1), AgGaSe2 (1.0 W m−<sup>1</sup> K−1) and BaGa4Se7 (0.56 W m−<sup>1</sup> K−1) [11,18]. With low thermal conductivity, the heat absorbed by the crystal from the pump pulse would accumulate because the pulse interval is shorter than the thermal-relaxation time. With the repetition rate increasing, the thermal effect would be more harmful [26]. The thermal diffusivity is 0.502 mm2 s−<sup>1</sup> for BGSe crystal along the *a* crystallographic axis [18], while the thermal diffusion time constant τ is about 376 ms, which is much longer than commonly used mid-infrared crystal, e.g., 5.7 ms for ZGP crystal. The prolonged accumulation of heat would lead to thermal lens effect and thermal dephasing which is unbeneficial to the parametric progress. Improving the cooling structure and a using narrow pump beam with high quality could rapidly reduce the thermal effects in BGSe crystal. The increase of the output energy of OPO at 250 Hz may be the joint result of the thermal focusing effect of Nd:YAG and the stabilization of the thermal effect in BGSe crystal.

The influence of the OPO cavity length to the input-output characteristics were studied at 4.06 μm with the fixed repetition rate at 100 Hz, as shown in Figure 4a. For different cavity lengths of 30 mm, 50 mm and 80 mm, the maximum output energies were 1.28 mJ/pulse, 1 mJ/pulse and 0.87 mJ/pulse, respectively. With the decrease of the cavity length, the transit period of the signal wave in the OPO cavity would decrease, which was beneficial to the interaction between the pump and the signal wave and led to the increase of the mid-infrared output energy.

**Figure 4.** (**a**) The input-output characteristics of the BGSe-based DP-SRO with different cavity lengths; (**b**) The conversion efficiency of the BGSe-based DP-SRO with different cavity lengths.

The OPO thresholds were measured to be 2.65 mJ/pulse, 3.38 mJ/pulse and 5.00 mJ/pulse for the cavity lengths of 30 mm, 50 mm and 80 mm, respectively. The experimental results verify the OPO theory established by S. J. Brosnan and R. L. Byer [22]. With the decrease of the cavity length, the instantaneous cavity net gain increased which leads to the decrease of threshold and the increase of conversion efficiency. The conversion efficiencies of the BGSe-OPO with different cavity lengths were shown in Figure 4b. The maximum conversion efficiency of 12.8% was obtained for the cavity length of 30 mm. For different cavity lengths of 30 mm, 50 mm and 80 mm, the slope efficiencies were measured to be 16.4%, 13.7% and 15.6%. The slope efficiency was measured to be almost constant for each cavity length, which means the back-conversion effect is negligible. With the increase of pump energy, the thermal effect caused by crystal absorption to the pump wave became more serious, causing the decrease of conversion efficiency.

By rotating the BGSe crystal, the tuning range of 3.62–4.10 μm was obtained in the BGSe-OPO. The theoretical wavelength-tuning curves versus the phase-matching angle of Type I BGSe-OPO is calculated based on the Sellmeier equations given in [27]:

$$m\_x^2 = 5.952953 + \frac{0.250172}{\lambda^2 - 0.081614} - 0.001709 \cdot \lambda^2 \tag{1}$$

$$m\_y^2 = 6.021794 + \frac{0.256951}{\lambda^2 - 0.079191} - 0.001925 \cdot \lambda^2 \tag{2}$$

$$m\_z^2 = 6.293976 + \frac{0.282648}{\lambda^2 - 0.094057} - 0.002579 \cdot \lambda^2 \tag{3}$$

The experimental results fit well with the theoretical calculation in Figure 5a. Figure 5b shows the tuning output characteristics of the BGSe-OPO. The tunable output decreased at both sides of the tuning curve due to the increasing pump loss from the Fresnel reflection. The abnormal valley of the tuning curve in Figure 5b around 4.34 μm was deduced to be related to the crystal impurities and defects.

**Figure 5.** (**a**) Comparison of the tuning curve between the experimental and the calculated results; (**b**) The tuning output characteristics of the BGSe-OPO.

Figure 6 shows the temporal pulse profiles of the pump wave and the mid-infrared wave, measured by Si photodiode (Thorlabs, DET025A/M, Newton, NJ, USA) and (HgCdZn) Te photodiode (Vigo, PCI-9, Ozarow, Mazowiecki, Poland), respectively. With the rapid decline of pump intensity at the trailing edge, the pulse width of the mid-infrared wave (13.6 ns) was measured to be shorter than the pump wave (20.0 ns) [28].

The spectrum of signal wave at normal incidence was recorded by a spectrometer, as shown in Figure 7. The Gaussian fitted curve has a full width at half maxima (FWHM) of 0.49 nm with a central wavelength of 1442.42 nm. According to the phase-match condition, the FWHM of generated mid-infrared is estimated to be 3.92 nm at 4061 nm, much narrower than the 2 μm pumped degenerate parametric conversion [21].

**Figure 6.** Comparison of the temporal pulse profiles of the pump wave and the mid–infrared wave.

**Figure 7.** The signal wave spectrum at normal incidence.

## **4. Discussion and Conclusions**

In this work, a 1064 nm laser pumping, tunable BGSe-OPO was demonstrated. Using a side-pumped electro-optical Q-switched Nd:YAG laser as the fundamental pump source, a tunable mid-infrared wave with high repetition rate up to 250 Hz was generated from

BGSe-OPO. DP-SRO configuration was utilized to improve the conversion efficiency and reduce the OPO threshold. The maximum average power of 250 mW was achieved at 4.06 μm, with the OPO cavity length of 30 mm. The influences of the repetition rate and the cavity length were studied experimentally. By rotating the BGSe crystal, tunable mid-infrared output from 3.42–4.73 μm was achieved. The pulse width of the generated mid-infrared wave was measured to be 13.6 ns and the linewidth at 4061 nm was estimated to be 3.92 nm.

The thermal effect of BGSe crystal was observed to be significant, especially in the high repetition rate OPO. Additionally, the accumulation of the thermal effect would result in the decrease of output power and LIDT with the increase of repetition rate, which was harmful to the parametric conversion process. Further optimizations in the pump beam quality, cooling method and cavity structure of DP-OPO would improve the output characteristics of BGSe-OPO.

The BGSe-OPO has advantages in generating the mid-infrared wave with widely tuning range, high power and high repetition rate. Furthermore, BGSe-OPO can be pumped with a 1064 nm laser which is low-cost and usually has a simple structure. BGSe-OPO has the advantages of both the tuning range of ZGP-OPO and the simple structure of PPLN-OPO, which are commonly used in commercial products. The system in this work will show great commercialization potential with the utilization of a commercial 1064 nm laser with a compact size.

**Author Contributions:** Conceptualization, R.Y. and J.Y. (Jiyong Yao); methodology, D.X.; validation, C.Y., K.C. and Y.H.; formal analysis, Y.W.; investigation, K.Z.; resources, J.Y. (Jiyong Yao); data curation, C.Y.; writing—original draft preparation, K.C.; writing—review and editing, Y.H.; visualization, K.C.; supervision, J.L.; project administration, D.X.; funding acquisition, D.X. and J.Y. (Jianquan Yao). All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Natural Science Foundation of China (NSFC), grant numbers U1837202, 62175182 and 62011540006.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

