A high-conversion-efficiency, low-threshold, quasi-continuous-wave optical parametric generator (OPG) based on a periodically poled lithium niobate (PPLN) crystal is presented. Pumped by an acousto-optically Q-switched 1064 nm Nd:YAG laser with a power output of 848 mW, the OPG generated an output power of 452 mW for the signal and the idle waves, achieving an internal conversion efficiency of 62.7% and a slope efficiency of 75.6%. To the best of our knowledge, this is the highest efficiency ever reported for single-pass, quasi-continuous-wave OPGs by using periodically poled crystals.
©2006 Optical Society of America
Optical parametric generation is an efficient way to obtain tunable coherent radiation in near-infrared and mid-infrared regions. Optical parametric oscillators (OPOs) and optical parametric generators (OPGs) based on periodically poled lithium niobate (PPLN) have been intensely investigated in the past few years. So far, the research works carried out on OPGs are much fewer than those on OPOs. This is partly because OPGs have higher thresholds and the lower conversion efficiencies. However, OPGs also have several advantages which make the applications of OPGs valuable. Firstly, an OPG requires no resonant cavity, which reduces the difficulty of adjustment and removes the coatings required for widely tunable and high-power operations. Secondly, for an OPG it is easy to inject a seed in the parametric generation process because of the omission of the cavity [1–2]. Therefore, it is possible to compensate for the broadened bandwidth introduced by the optical parametric generation. Thirdly, an OPG is a promising optical source with a widely tunable bandwidth. Broadband optical radiation in a PPLN crystal has been generated by using a non-collinear quasi-phase-matching scheme and an elliptical pump beam .
Compared with conventional birefringent phase-matching (BPM) techniques, quasi-phase-matching (QPM) technique has several advantages. The most important one is the utilization of the largest nonlinear coefficient which makes all kinds of nonlinear processes involved with QPM more efficient than the conventional methods. Among various QPM crystals, PPLN crystals are the most widely used ones. Thus, PPLN-based OPGs are important devices for achieving widely tunable, highly efficient optical sources. M. Rahm et al reported an OPG with a 45% conversion efficiency by using a 55 mm long PPLN pumped by a 10 kHz nanosecond single-mode Nd: YVO4 laser . Mark J. Missey et al obtained a conversion efficiency of 62% (based on a measured pump depletion) in a pulsed OPG. However, the repetition rate of their system was only 30 Hz . We previously demonstrated a PPLN OPG with a conversion efficiency of 32.9% at a repetition rate of 22.6 kHz and with a pulse width of 12 ns . Using a 38.7 mm-long PPLN crystal and a 6.58 W input at 1064 nm, we also achieved a 1.9 W OPG output at 50 kHz .
In this paper, we report an experimental demonstration of a high-efficiency, low-threshold, high-repetition-rate PPLN OPG. A conversion efficiency (defined as the ratio of the total OPG output to the 1064nm pump input) of 62.7% has been achieved. When the Fresnel reflection losses are taken into account, the internal conversion efficiency would be higher than 62.7%. To the best of our knowledge, this is the highest conversion efficiency ever reported for single-pass, high-repetition, quasi-continuous-wave PPLN OPGs.
In the process frequently used for QPM OPGs, the only nonlinear coefficient involved is d33 which has the largest value. All of the waves (pump, signal and idle) propagating in a PPLN crystal are extraordinary waves and correlate to each other through the coupled-mode equations. The analytical solutions have been developed using the small signal approximation and are given in the literature [8–10].
To discuss the factors that influence the conversion efficiency of the optical parametric generation, we need to theoretically calculate the gain curve of the optical parametric generation process with different PPLN periods, at different temperatures, and especially in different pump intensities. The expression of OPG gain was already derived by Myers  and can be written as
where ωs, ωi are the angular frequencies of the signal and idle waves; dQ is the effective nonlinear coefficient; Ip is the pump intensity, ns, ni, np are the refraction indexes of the signal, idle, and pump waves respectively; ε 0 is the vacuum electric susceptibility; c is the velocity of light in vacuum; L is the length of the PPLN crystal; ΔkQ is wave-vector mismatch. It can be seen from Eq. (1) that the gain is proportional to the intensity of pump wave. The pump intensity can be written as
where Pa is the average pump power; f is the repetition rate of the pump, τ is the duration of each pump pulse; and ω is the waist of the pump beam inside the crystal.
In order to estimate the bandwidth of the OPG, we adopted the transfer matrix method as reported in Ref.  and  to calculate the amplification coefficient in the OPG. Using Eqs.(4)-(6) in Ref. , the amplification coefficient as a function of wavelength can be calculated. Thus the bandwidth of the OPG could be estimated from the calculation.
3. Experiments and results
Figure 1 shows a schematic diagram of the PPLN OPG used in our experiments. A diode end-pumped and acousto-optically Q-switched Nd:YAG laser was built and used as the pump source. At a driving current of 33.6 A, the laser diode generated a CW output of 15.36 W at 808 nm. A power delivery fiber with a core diameter of 400 μm and a numerical aperture of 0.22 was used to deliver the light from the laser diode to an optical coupler. The coupler consisting of a pair of convex lens and with a compression ratio of 1:1 coupled the 808 nm pump radiation into a Φ3×8 mm Nd: YAG rod. The input facet of the Nd: YAG rod was coated with antireflection coatings at 808 nm and high reflection coatings at 1064 nm. An active acousto-optical Q-switch with repetition rates tunable from 1 kHz to 50 kHz was placed in the cavity to ensure the quasi-continuous-wave lasing operation. A Brewster plate was inserted into the cavity in order to achieve linearly polarized operation at 1064 nm. The output coupler of the 1064 nm pump laser was a plane mirror with a transmission of 30% at 1064 nm. When the output power of the 1064 nm laser was 700 mW, we obtained a near-diffraction-limited beam with a beam quality factor M2 < 1.1. The pulse duration was measured to be approximately 9 ns (FWHM). A convex lens with a focal length of 100 mm was placed after the 1064 nm pump output coupler so that sufficient pump intensity could be achieved. The beam waist diameter at the focusing point (measured without the presence of the crystal) was about 55 μm and the pump waist inside the crystal was calculated to be 76.4 μm. The PPLN crystal used in our experiments was a 48.87 mm×7.7 mm×0.5 mm multi-grating PPLN crystal containing 13 sections with periodically poled periods ranging from 26 μm to 32 μm in a 0.5 μm step. Both the facets of the crystal were coated with antireflection coatings for three wavelength ranges, that is, 1064 nm, 1350-1700 nm and 2846-5027 nm. We placed the PPLN crystal in a heating oven to achieve temperature tuning. The temperature of the crystal could be precisely tuned from room temperature to 250°C with an accuracy of ±0.1°C.
In order to increase the conversion efficiency, we used the polarized output of the Nd: YAG pump by adding the Brewster plate. We found that the Q-switched pulse shape was significantly improved compared with the unpolarized operation and the trailing problem was successfully solved. According to Eq. (1), the gain in an OPG is proportional to the pump intensity and the square of the PPLN crystal length. In our experiments, sufficient pump intensity was achieved by focusing the pump beam. We found that the system was optimal when the repetition rate of the Q switch was set at 4.1 kHz. When the section with a poled period of 31 μm was used and the crystal temperature was set at 168°C, a signal wave with a central wavelength of 1920 nm and an idle wave at 2387 nm were simultaneously generated. The OPG powers under different pump inputs were measured by using a power meter (Molectron EPM1000) and are shown in Fig. 2. At a pump power of 848 mW at 1064 nm, a total output power of 452 mW was achieved. It should be noted that the detector of the power meter was placed after a filter which was coated with high-reflection coatings for 1064 nm and 532 nm. Since no high transmission coatings for the signal and idle waves were applied to the filter, a 15% could be involved in the power measurements. Therefore, when taking the transmission loss into account, the actual output was approximately 531.8 mW, and the crystal internal conversion efficiency and the slope efficiency would be 62.7% and 75.6% respectively. Since the antireflection coatings of the PPLN crystal only covered the 1350-1700 nm and 2846-5027 nm ranges, losses also existed for the 1920 nm signal wave and its corresponding idle wave at the crystal facets. When this condition is considered, the actual conversion efficiency should be higher. Furthermore, we did not consider the Fresnel reflection losses at the crystal facets. The beam quality factor of the OPG output was measured as M2=1.76 at a pump power of 488 mW, indicating that the beam quality deteriorated as compared with the 1064 nm pump beam. We observed the so-called pump depletion phenomenon (approximately 77%) in the experiments, as in Fig. 3. The obvious pump depletion corresponds to the high-conversion process.
Compared with OPOs, the conventional OPGs normally show high thresholds because of the absence of resonant cavities. However, the OPG threshold in our experiments was still acceptable and even comparable with some OPO systems. The threshold value was measured to be 169 mW for a repetition rate of 4.1 kHz. At this threshold, the energy of each pump pulse was 41.2 μJ.
Figure 4 shows a spectrum of the OPG output measured when the maximum conversion efficiency was achieved. Both the signal and idle waves had bandwidths as wide as 40 nm. The broad bandwidth is also due to the absence of a resonant cavity that can select the modes of oscillation and narrow the bandwidth. In fact, broadband infrared sources can be used in different applications such as medical imaging, pollution detection etc. By using the formulas in Ref. , we calculated the gain of the OPG as a function of wavelength. The calculated results are shown in the inset of Fig. 4. It can be seen that the theoretical curve agrees well with the experimental observation. We also calculated the bandwidth by using Eq. (1), but the calculated results showed dissatisfied agreement with the experimental observation. In general, the higher pump intensity can produce broader bandwidth.
In summary, we have demonstrated, to the best of our knowledge, the highest conversion efficiency operation of a low-threshold, high-repetition-rate PPLN OPG. The OPG was pumped by a diode-end-pumped acousto-optically Q-switched 1064 nm Nd:YAG laser. At a pump input of 848mW and a repetition-rate of 4.1 kHz (pulse duration 9 ns), an internal conversion efficiency of 62.7% and a slope efficiency of 75.6% were achieved by using the 31μm-period section of the multi-grating PPLN (set at 168°C). The threshold of the OPG was measured as 41.2 μJ. The relationship between the OPG bandwidth and the pump intensity has been investigated both theoretically and experimentally. Our results show that the OPG is a useful and simple way to generate broadband mid-infrared radiations.
We would like to thank Mr. Dingwen Zhou and Miss Huimin Zhang for precisely measuring the transmission of the filter used in the experiments. This work was supported in part by the National Natural Science Foundation of China (Grant Nos. 10474071 and 60578054).
References and links
1. U. Bader, T. Mattern, T. Bauer, J. Bartschke, M. Rahm, A. Borsutzky, and R. Wallenstein, “Pulsed nanosecond optical parametric generator based on periodically poled lithium niobate,“ Opt. Commun. 217, 375–380 (2003). [CrossRef]
2. B. Kohler, U. Bader, A. Nebel, J. P. Meyn, and R. Wallenstein, “A 9.5-W 82-MHz-repetition-rate picosecond optical parametric generator with cw diode laser injection seeding,“ Appl. Phys. B 75, 31–34 (2002). [CrossRef]
3. S. M. Russell, P. E. Powers, M. J. Missey, and K. L. Schepler, “Broadband mid-infrared generation with two-dimensional quasi-phase-matched structures,“ IEEE. J. of Quantum Electron. 37, 877–887 (2001). [CrossRef]
4. M. Rahm, U. Bader, G. Anstett, J. P. Meyn, R. Wallenstein, and A. Borsutzky, “Pulse-to-pulse wavelength tuning of an injection seeded nanosecond optical parametric generator with 10 kHz repetition rate,“ Appl. Phys. B 75, 47–51 (2002). [CrossRef]
5. V. D. Mark, J. Missey, Peter E. Powers, and Kenneth L. Schepler, “Periodically poledlithium niobate monolithic nanosecond optical parametricoscillators and generators,“ Opt. Lett. 24, 1227–1229 (1999). [CrossRef]
6. B. G. Zhang, J. Q. Yao, X. Ding, H. Zhang, P. Wang, D. G. Xu, G. J. Yu, and F. Zhang, “Low-threshold, high-efficiency, high-repetition-rate optical parametric generator based on periodically poled LiNbO3,“ Chin. Phys. 13, 364–368 (2004). [CrossRef]
7. B. G. Zhang, J. Q. Yao, Y. Lu, D. G. Xu, J. Feng, T. L. Zhang, X. Zhao, P. Wang, and K. X. Xu, “High-average-power nanosecond quasi-phase-matched single-pass optical parametric generator in periodically poled lithium niobate,“ Chin. Phys. Lett. 22, 1691–1693 (2005). [CrossRef]
8. C. Q. Xu, H. Okayama, and M. Kawahara, “Optical frequency conversions in nonlinear medium with periodically modulated linear and nonlinear optical parameters,“ IEEE. J. of Quantum Electron. 31, 981–987 (1995). [CrossRef]
9. Y. Zhang and B. Y. Gu, “Optimal design of aperiodically poled lithium niobate crystals for multiple wavelengths parametric amplification,“ Opt. Commun. 192, 417–425 (2001). [CrossRef]
10. L. E. Myers, R. C. Eckardt, M. M. Fejer, R. L. Byer, W. R. Bosenberg, and J. W. Pierce, “Quasi-phase-matched optical parametric oscillators in bulk periodically poled LiNbO3,“ J. Opt. Soc. Am. B 12, 2102–2116 (1995). [CrossRef]