Open Access
Add to BibSonomy
Abstract
We developed a visible-red to near-infrared wavelength tunable all-solid-state laser system utilizing an optical parametric generation process in a MgO doped PPLN crystal pumped at 532 nm by an amplified and frequency doubled picosecond passively Q-switched Nd:YVO4 microchip laser. A broad bandwidth, tuneable over 300 nm between 710 nm to 1015 nm, is accessible. Depending on the green pump light pulse energy, pulses with durations down to 69 ps as well as pulses with energies above 2 µJ were achieved with kHz repetition rates.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Monolithic integrated passively Q-switched solid-state microchip lasers are well established and robust devices, also very suitable for wafer scale mass production [1]. Particularly in combination with a semiconductor saturable absorber mirror (SESAM), they close the gap between nanosecond Q-switched lasers and a few tens of picosecond mode-locked lasers. Their benefit due to their short cavity length is to generate single frequency sub-ns to sub-100 ps pulses with several tens to hundreds of nJ pulse energy [1–3]. With an additional amplifier stage in a master oscillator power amplifier (MOPA) configuration, these compact all-solid-state devices are interesting for applications such as laser machining or LIDAR [4,5]. Combining these devices with nonlinear crystals for higher harmonic generation or optical parametric processes , a broad wavelength range is accessible [6–8]. In most cases, the generated wavelengths are in the mid IR range and hence longer than the typical Nd:YVO$_4$, Nd:YAG or Yb:YAG microchip seed laser wavelength [9–12]. In contrast, we developed a sub-200 ps tuneable laser system in the visible-red to near-infrared (NIR) wavelength range based on an amplified and frequency doubled (second harmonic generation - SHG) Nd:YVO$_4$ microchip laser and an optical parametric generation (OPG) process in a MgO:PPLN crystal pumped at 532 nm. The system was designed for short pulse excitation for photoluminescence spectroscopy and for optically pumping GaAs/InGaAs semiconductor laser (OPSL) structures with design wavelengths around 1030 nm. The system parameters to achieve were a wavelength tuning range of at least 800 nm to 1000 nm, pulses with durations below 200 ps and energies of at least 150 nJ. The broad wavelength range was necessary to allow the excitation of barrier states as well as the discrete energy states of the GaAs/InGaAs quantum well for barrier and in-well pumping of OPSLs.
2. Experimental setup
The system depicted in figure 1 is based on four stages: a fully integrated oscillator module emitting at 1064 nm, a double pass amplifier and a single pass SHG stage to pump a magnesium oxide doped periodically poled LiNbO$_3$ crystal (MgO:PPLN) at 532 nm. The self-built oscillator module (see Fig. 2) combines a free space broad area single emitter laser diode as the pump source at 808 nm , pump beam shaping optics for the fast and slow axis of the laser diode and a passively Q-switched microchip laser as well as a thermoelectric cooler and heat sensor on a total footprint of 20 mm x 25 mm. The pump diode is a 90 µm broad area single emitter chip on a submount emitting up to 4 W of cw output power (II-VI SES4-808C-01). The linewidth is about 2 nm. The beam quality factor M$^2$ was determined to $1.2$ for the fast axis and to $20$ for the slow axis, based on an approximation after a measurement of the pump spot size with the knife-edge method. Compared to a fiber-coupled pump diode with its randomly distributed light polarization, the linearly polarized pump radiation of the single emitter will increase the pump absorption efficiency as the absorption cross sections of Nd:YVO$_4$ differ significantly for the c- and a-axis of the crystal.
The microchip laser itself is based on a 200 µm thin commercially available Nd:YVO$_4$ crystal, emitting at 1064 nm, with an a-cut and a doping concentration of 2 at% to achieve high pump absorption. A dielectric coating serves as the output coupler on the top side with a transmission of 10 %. A half-wave plate between the laser diode and the microchip laser ensured the correct polarization orientation for highest pump light absorption in the Nd:YVO$_4$ crystal. For passive Q-switching, the crystal is glued by a sol-gel based process on a high quality SESAM with a modulation depth of 12.6 %, a recovery time of about 1 ns and non-saturable losses of 0.2 %, manufactured by FBH. The SESAM was identical to the one we used in a previous work to demonstrate repetition rates up to 1 MHz with a wedged crystal approach [2]. However, in this work we limited the repetition rate to below 200 kHz since we observed changes in the beam characteristics due to changes of the pump spot size of the single emitter on the microchip for higher pump currents. In all cases the beam profile was Gaussian like. The amplifier stage consisted of a double pass in an 8.2 mm long end pumped Nd:YVO$_4$ crystal with an a-cut and anti-reflection coatings of the end facets for both the pump and the laser emission at 808 nm and 1064 nm. A low doping concentration of 0.4 at% ensured a homogeneous pump light distribution. The crystal was cw pumped by a fiber coupled and temperature controlled laser diode with 20 W of output power. To avoid damages of the pump diode, two HT808 / HR1064 dichroic mirrors were placed between the pump fiber and the crystal. The laser beam from the microchip laser was collimated with an 80 mm lens and focused with a 150 mm lens into the crystal of the amplifier to achieve a focused spot size that was slightly smaller than that of the amplifier pump diode. The latter was fiber coupled with a core diameter of 105 µm and was focused into the amplifier crystal via a 1:1 imaging optics. For all following results, the amplifier pump power was set to 20 W. Two Faraday isolators were used to protect the oscillator from backreflections and to separate the non-amplified seed from the amplified beam via the polarization state.
Fig. 1. Setup of the laser system. $\mu$L: microchip laser seed module, L1: 80 mm lens for collimation, FI: Faraday isolator, $\lambda /2$-WP: half-wave-plate, L2: 150 mm lens, LD: laser diode, L3: 200 mm lens, DM: dichroic mirror, L4: 200 mm lens, L5: 150 mm lens, BD: beam dump.
Fig. 2. Photo of the self-built oscillator module. 1: single emitter on submount, 2: fast-axis collimation lens, 3: half-wave plate, 4: slow-axis collimation lens, 5: focusing lens, 6: Nd:YVO$_4$ microchip laser, 7: temperature sensor, 8: AlN base plate on thermoelectric cooler.
The second harmonic is generated in a 10 mm long lithium triborate (LBO) crystal at a critical phase-matching angle of 11.6°, fine tuned for highest average green output power by heating the crystal moderately to around 38 °C in a self-built temperature controlled oven. We focused the amplified seed beam with a 200 mm lens to generate a focused beam waist diameter of about 130 µm at the center of the LBO crystal. The polarization of the fundamental wavelength was adapted for optimum frequency conversion with a half-wave plate. Finally, the 532 nm beam was focused into multi grating MgO:PPLN crystal (Covesion Ltd.) via a 150 mm lens, which generated a focused spot diameter of 80 µm. We chose a crystal length of 20 mm based on the pump pulse parameters and the requirements described in the introduction regarding output pulse duration and energy but also regarding high conversion efficiency and narrow output bandwidth. The MgO:PPLN crystal had five gratings with periods of 6.81 µm, 7.1 µm, 7.4 µm, 7.71 µm and 8.03 µm for quasi-phase matching and a broadband AR coating for the pump at 532 nm and the whole range of the signal and idler emission. Each periodically poled grating is separated by 0.2 mm wide regions of unpoled material. For wavelength tuning at a given grating period, the MgO:PPLN crystal was heated in an oven which was mounted to a linear stage for changing the gratings easily. The oven could heat the MgO:PPLN crystal to temperatures between 30 °C to 200 °C. The remaining pump and idler light at the output of the MgO:PPLN crystal was spectrally filtered out by a combination of optical long and short pass dichroic mirrors. Pulse durations were measured with a 12 GHz oscilloscope (Agilent DSO81204B) and a fast fiber connected photodetector (Newport Model 1414). In general, the pulse energies were derived from a measurement of the average power using an integrating sphere power sensor with Si photodiode (Thorlabs S142C), which shows strong wavelength dependency of the responsivity, as well as with a thermal power sensor (Thorlabs S405C) with a very flat responsivity curve in the studied wavelength range. Both methods gave very similar results.
3. Results
The microchip laser emitted pulses with a length of 187 ps and energy of about 300 nJ over a large repetition rate range, here up to 170 kHz, as expected [2]. However, since guiding the seed laser beam through the two isolators causes some losses, only about 180 nJ entered the amplifier (see Fig. 3(a), black squares). At repetition rates as low as 11.6 kHz we achieved up to 23.4 µJ of pulse energy at 1064 nm (see Fig. 3(a), red dots) which corresponds to an amplification by over 23 dB in this double pass configuration. With increased repetition rate the extracted pulse energy decreased as the rate of pulse amplification by stimulated emission rises and depletes the level of inversion gradually. But even at about 170 kHz, the seed laser pulse was amplified up to 8.2 µJ or by about 16.7 dB. Over the complete tested range of repetition rates, the shape und hence the pulse duration of the amplified pulses were unaffected by the amplification process.
Fig. 3. (a) Pulse energy over repetition rate for the amplifier input (black), the amplified seed (red), the SHG output (green) and the OPG output at 797 nm (blue). (b) Output pulse energy of the OPG signal path (black dots) and the pulse duration (FWHM) at 797 nm (OPG, red) and 532 nm (SHG, green) over green pump pulse energy at 532 nm.
For the SHG, pulse energies from 9 µJ at 11.6 kHz to about 2 µJ at 170 kHz were achieved with a conversion efficiency decreasing from 45 % to 20 % with increasing repetition rate. This corresponded to pump peak powers of 123 kW to 43.9 kW at 1064 nm. For an exemplary repetition rate of 44.5 kHz, we achieved a slope efficiency of close to 42 %. Even though MgO doping of PPLN can decrease the effect of green-induced infrared absorption (GRIIRA) to a negligible level [13], care has still to be taken for potential damages at high pump intensities. Nevertheless, even at green pump light fluences of up to 179 mJ/cm2, we did not observe any photodarkening effects or irreversible damage of the crystal after months of daily operation. The multiple gratings of the MgO:PPLN crystal allowed to generate picosecond laser pulses over a broad wavelength range of over 300 nm from 710 nm to 1015 nm. Because the quasi-phase-matching condition could be adjusted by heating the MgO:PPLN crystal in the oven and the gratings overlapped spectrally, all wavelengths within the tuning range were accessible as depicted in Fig. 4(a). A slight drawback was the increased phase-matching bandwidth of the 6.81 µm grating that resulted in an increased spectral width of the output emission for a given temperature (see Fig. 4(b). Also the tuning sensitivity was increased. Altering the temperature by 40 °C shifted the peak wavelength by 71 nm. Within the same temperature range, the grating with the next longer period of 7.1 µm only shifted the peak wavelength by 22 nm and the emission bandwidth decreased further for increasing grating periods. The broadening of the emission bandwidth is a common observation when the wavelengths of the signal and idler converge towards the degeneracy wavelength, which is 1064 nm in our setup [11]. It depends on the specific application whether the spectral width of the emission is acceptable or not. In our case of exciting semiconductor laser structures, the spectral width was narrow enough to excite the different discrete energy states in the quantum wells. Regarding the polarization state of the signal light, the type-0 phase matching due to the periodic poling in the lithium niobate ensured that all the involved light waves had the same polarization state. We verified this by measuring the power of the signal path before and after a broad band antireflection coated polarizing beam splitter cube and derived a content of above 98 % of e-polarization, following the polarization state of the green pump light.
Fig. 4. (a) Normalized spectra over the observed wavelength range and maximum pulse energy at 44.1 kHz repetition rate over wavelength. (b) Spectral width (FWHM) of the measured spectra in 4(a). The highlighted background indicates the used grating.
In Fig. 3(a) we show the pulse energy as a function of the repetition rate for an exemplary output wavelength of 797 nm (blue triangles). At 11.6 kHz we achieved above 2.1 µJ and still more than 250 nJ at 170 kHz with conversion efficiencies of 23.3 % and 12.5 %, respectively. As an example, at the same wavelength and at a repetition rate of 54 kHz, we achieved a slope efficiency of green pump light pulse energy to the OPG pulse energy of about 26 % (see Fig. 3(b), black dashed line). For a fixed repetition rate of 44.1 kHz, we determined the dependence of the output parameters on the output wavelength. We obtained the highest energy and average output power close to the center of the emission bandwidth at about 850 nm and a decrease towards the edges (see Fig. 4(a), blue data points). For the slope in Fig. 3(b), we varied the green pump pulse energy of the PPLN crystal by attenuating the fundamental pump pulse energy at 1064 nm. In principle, this might have affected parameters like the pulse duration of the green light. However, we measured the 532 nm pulse duration after the SHG and this was only gradually increasing over the measured pump range. One can, thus, assume that the OPG pulse length is not affected by these minimal changes (Fig. 3(b), green rhombus). As observed before, the high gain or exponential gain regime of the OPG process results in a pulse shortening compared to the pump pulse duration [6,10,14]. For the case of negligible pump depletion, this can be approximated by [15]
limited by the walk-off and the dispersion length. With the signal photon gain path length L, the pump pulse duration $\tau _p$, the phase-matched gain coefficient $\Gamma = \sqrt {2\omega _s\omega _i \left |d_{\text {eff}}\right |^2 I_p / \epsilon _0 n_s n_i n_p c^3}$, $c$ the vacuum speed of light, $\epsilon _0$ the vacuum permittivity, $n$ wavelength dependend refractive index for the the involved wavelengths, $I_p$ the undepleted pump intensity and the effective nonlinear coefficient $d_{\text {eff}}$ of the PPLN crystal. Equation 1 can describe our observation of an OPG output pulse as short as 69 ps for 797 nm just above the OPG threshold as plotted in figure 5(a) (blue line). The pulse width is close to half the width of the green light pump pulse and hence $\Gamma \cdot L \approx 8$ for this pump energy. However, as depicted in figure 5(b), the green pump light pulses, measured behind the PPLN crystal, experienced strong pump depletion resulting in a strong dip at the center (red) or an almost complete extinction (blue) of the pump pulse. Therefore the model and approximation of negligible pump depletion is not valid anymore. With increasing pump depletion and hence increasing conversion efficiency, the OPG pulse width increased, too (see Fig. 3(b), red triangles). This effect was observed before, as only the part of the pump pulse above threshold takes part in the OPG process [11].Fig. 5. (a) Oscilloscope traces of the seed pulse and the minimum pulse duration of the the SHG and OPG signal at 532 nm and 797 nm, respectively. (b) Oscilloscope traces of the SHG pulses after the PPLN crystal highlighting the increasing effect of pump depletion in the OPG process.
For utilizing the system as a light source for spectroscopic applications, the stability of the pulse energy from pulse to pulse is an important parameter. Due to the pulse build up from spontaneous emission, slight fluctuations of the pulse energy and timing jitter are inherent characteristics in a passively Q-switched microchip laser. Methods like self-injection seeding [16–18] are suitable to reduce or avoid these fluctuations. Keeping this in mind, we measured the pulse energy at 950 nm over 6 h (see Fig. 6(a)). As the used energy meter (Ophir Centauri + PE9-ES-C) was only capable to detect every single pulse for a repetition rate below 25 kHz, a slightly lower pulse repetition rate of 18.7 kHz was chosen. The mean value was 1.36 µJ with a standard deviation of 8.46 nJ or 0.62 %. Additionally, we did a long term stability test beyond 1000 h of operation for the microchip laser only (see Fig. 6(b)) to investigate the reliability of the glue layer between the laser crystal with the SESAM and the reliability of the SESAM itself under operational conditions. In this long term test, we measured the average output power and the sampling rate was every 30 s to see potential trends. The microchip was mounted on a passive heat sink and was pumped by a fiber coupled laser diode at 808 nm. The fiber had a core diameter of 105 µm and we used 2:1 imaging optics to generate a focused spot size of about 60 µm. The cw pump power was set to a rather high value of 0.5 W to accelerate potential degradation effects like photodarkening or irreversible damages. The setup in this test was not shielded against mechanical vibrations or enviromental fluctuations of the room temperature in the lab for the benefit of changing and characterizing a number of microchips easily. Therefore, one can identify the day and night as well as the weekend cycle in the average output power in figure 6(b). However, we did not observe any significant changes of the operation parameters that would result from photodarkening or damages of the glue layer or the SESAM.
Fig. 6. (a) Pulse energy of the OPG system for over six hours of operation at 950 nm and 18.7 kHz. (b) Long term testing of the used microchip laser over 1000 hours of operation. Pump power at 808 nm was about 0.5 W resulting in a repetition rate of about 502 kHz.
4. Conclusion
In this work, we demonstrated an all-solid-state, wavelength tuneable picosecond laser system for the visible-red to near-infrared band based on an amplified and frequency converted passively Q-switched microchip laser. High pulse energies above 1 µJ with pulse durations of around 145 ps over the whole tuning range from 710 nm to 1015 nm and above 2 µJ in maximum were achieved. As the pulse duration was sensitive to the pump power, durations between 69 ps to 145 ps were measured over the whole tuning range. The strongly application driven system was designed for excitation of semiconductor laser structures based on the GaAs/InGaAs material systems and the obtained results will be published elsewhere. The target parameters regarding pulse duration, pulse energy and wavelength tuning range, defined in the introduction part, were in some cases significantly exceeded. Due to the very similar wavelength tuning range, one might compare the system with typical titanium sapphire picosecond mode-locked lasers. While the pulse duration might be several times longer in our system, the high pulse energy from the microchip seed laser does not need a sophisticated power amplifier design. Moreover, a high speed pulse picker is not required either since a continuous tuning of the repetition rate in the kHz range is possible by adjusting the microchip pump power. This makes our solution attractive for applications where the constraints on pulse duration are less strong. Even shorter pulses appear possible with the usage of compression techniques [19,20]. As an outlook, adding an extra SHG stage for entering the UV-blue-green wavelength range for pumping wide band gap semiconductor materials like GaN/InGaN will be very attractive.
Funding
Bundesministerium für Wirtschaft und Energie (16KN053064).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. J. J. Zayhowski, “Microchip lasers,” Opt. Mater. 11(2-3), 255–267 (1999). [CrossRef]
2. A. Marianovich, S. Spiekermann, M. Brendel, P. Wessels, J. Neumann, M. Weyers, and D. Kracht, “Wedged Nd:YVO4 crystal for wavelength tuning of monolithic passively Q-switched picosecond microchip lasers,” Opt. Express 29(13), 19790–19795 (2021). [CrossRef]
3. S. Forget, F. Balembois, G. Lucas-Leclin, and P. Georges, “Picosecond laser source at 1 MHz with continuous tunability in the visible red band,” Opt. Commun. 220(1-3), 187–192 (2003). [CrossRef]
4. M. A. Albota, R. M. Heinrichs, D. G. Kocher, D. G. Fouche, B. E. Player, M. E. O’Brien, B. F. Aull, J. J. Zayhowski, J. Mooney, B. C. Willard, and R. R. Carlson, “Three-dimensional imaging laser radar with a photon-counting avalanche photodiode array and microchip laser,” Appl. Opt. 41(36), 7671–7678 (2002). [CrossRef]
5. J. Degnan, D. Wells, R. Machan, and E. Leventhal, “Second generation airborne 3D imaging lidars based on photon counting, Advanced Photon Counting Techniques II, vol. 6771W. Becker, ed. International Society for Optics and Photonics (SPIE, 2007), pp. 117–123.
6. J. J. Zayhowski, “Periodically poled lithium niobate optical parametric amplifiers pumped by high-power passively Q-switched microchip lasers,” Opt. Lett. 22(3), 169–171 (1997). [CrossRef]
7. A. Marianovich, S. Spiekermann, M. Brendel, P. Weßels, J. Neumann, M. Weyers, and D. Kracht, “Picosecond Optical Parametric Generation in the Red to Near Infrared Band with a MgO:PPLN Crystal Pumped by an Amplified Passively Q-switched Microchip Laser,” in in Laser Congress 2020 (ASSL, LAC) (Optical Society of America, 2020), p. AF3A.6.
8. A. P. Piskarskas, “Optical parametric tunable, generators: Powerful & ultrafast,” Opt. Photonics News 8(7), 24 (1997). [CrossRef]
9. H. Ishizuki and T. Taira, “High-gain mid-infrared optical-parametric generation pumped by microchip laser,” Opt. Express 24(2), 1046–1052 (2016). [CrossRef]
10. Y.-C. Chiu, Y.-C. Huang, and C.-H. Chen, “Parametric laser pulse shortening,” Opt. Lett. 39(16), 4792–4795 (2014). [CrossRef]
11. A.-C. Chiang, T.-D. Wang, Y.-Y. Lin, C.-W. Lau, Y.-H. Chen, B.-C. Wong, Y.-C. Huang, J.-T. Shy, Y.-P. Lan, Y.-F. Chen, and P.-H. Tsao, “Pulsed optical parametric generation, amplification, and oscillation in monolithic periodically poled lithium niobate crystals,” IEEE J. Quantum Electron. 40(6), 791–799 (2004). [CrossRef]
12. G. J. Spühler, R. Paschotta, M. P. Kullberg, M. Graf, M. Moser, U. Keller, L. R. Brovelli, C. Harder, E. Mix, and G. Huber, “Passively Q-switched Yb:YAG microchip laser using a semiconductor saturable absorber mirror,” in Advanced Solid State Lasers (Optical Society of America, 1999), p. TuC2.
13. Y. Furukawa, K. Kitamura, A. Alexandrovski, R. K. Route, M. M. Fejer, and G. Foulon, “Green-induced infrared absorption in MgO doped LiNbO3,” Appl. Phys. Lett. 78(14), 1970–1972 (2001). [CrossRef]
14. A. C. Chiang, Y. C. Huang, Y. W. Fang, and Y. H. Chen, “Compact, 220-ps visible laser employing single-pass, cascaded frequency conversion in monolithic periodically poled lithium niobate,” Opt. Lett. 26(2), 66–68 (2001). [CrossRef]
15. R. L. Byer, “9 - optical parametric oscillators,” in Quantum Electronics - Vol. 1 - Nonlinear Optics - Part B, H. Rabin and C. Tang, eds. (Academic Press, 1975), pp. 587–702.
16. A. Steinmetz, D. Nodop, A. Martin, J. Limpert, and A. Tünnermann, “Reduction of timing jitter in passively Q-switched microchip lasers using self-injection seeding,” Opt. Lett. 35(17), 2885–2887 (2010). [CrossRef]
17. M. Hao, S. Wang, Y. Wang, G. Feng, and S. Zhou, “Timing jitter effect in microchip laser and its reduction by backward reflection in fiber delay line,” Opt. Eng. 58(05), 1–6 (2019). [CrossRef]
18. J. A. Brooks, G. M. Bonner, A. J. Kemp, K. Oakes, and D. J. M. Stothard, “Pathways to Reducing Jitter in Q-Switched and Cavity-Dumped 2 µm Lasers, in Laser Congress 2017 (ASSL, LAC) (Optical Society of America, 2017JTu2A.25–00
19. R. Diao, Z. Liu, F. Niu, A. Wang, T. Taira, and Z. Zhang, “Compressed 6 ps pulse in nonlinear amplification of a Q-switched microchip laser,” Laser Phys. 27(2), 025102 (2017). [CrossRef]
20. A. Steinmetz, F. Jansen, F. Stutzki, R. Lehneis, J. Limpert, and A. Tünnermann, “Sub-5-ps, multimegawatt peak-power pulses from a fiber-amplified and optically compressed passively Q-switched microchip laser,” Opt. Lett. 37(13), 2550–2552 (2012). [CrossRef]
