We report on the implementation of an all-solid-state optical parametric oscillator (OPO) laser system, pumped by a fiber laser, and extended by intra-cavity sum frequency generation (SFG) to provide tunable radiation with output powers well beyond 1 W in the visible regime between 605 and 616 nm. We use periodically poled sections for quasi phase-matched OPO and SFG processes, implemented on a single MgO:PPLN crystal. A Pound-Drever-Hall frequency stabilization reduces the laser linewidth to the range of 100 kHz (FWHM), determined by measurements of spectral hole burning in a rare-earth ion doped crystal as well as analysis of side-of-fringe transmission in a low finesse Fabry-Perot resonator.
© 2014 Optical Society of America
Tunable, high-power, continuous-wave (cw) laser sources are a prerequisite for innumerous applications in optics. Optical parametric oscillators (OPOs) provide a powerful approach to implement such light sources. In an OPO, a strong pump wave drives nonlinear parametric generation of a signal wave and an idler wave. Spectral tuning of the signal and idler wave is accomplished by phase-matching. In recent years, the development of efficient cw-OPOs (see e.g . and Refs. therein) was significantly pushed forward by the commercial availability of periodically poled crystals [2,3], e.g. periodically poled lithium niobate (PPLN). These crystals combine large nonlinear optical coefficients with quasi phase-matched operation at large interaction length. To obtain large conversion efficiencies, typically PPLN crystals in cw-OPOs are pumped by high-power infrared lasers. Thus, the generated signal and idler wavelength from the OPO will be far in the infrared spectral range.
However, many applications require visible, and more specifically orange through red radiation. There are several ways to achieve this goal. Most demonstrations of visible cw-OPOs have used a pump source at a shorter wavelength, primarily at λP = 532 nm . Direct generation of visible light in the signal wave of the OPO has proven to be challenging due to limitations of the nonlinear materials under exposure to high visible pump and circulating powers. A more successful approach has proven to be intra-cavity frequency-doubling of a resonant infrared idler wave . However although this approach has been demonstrated with excellent frequency stability, the powers and efficiencies were relatively low (P = 30 mW and λ = 604 nm output at PP = 3.4 W and λP = 532 nm pump) due to the requirement to use critically-phase-matching for frequency-doubling. By comparison, our approach - pumping at a near-infrared wavelength, and applying additional intra-cavity sum frequency generation (SFG) of the generated signal (or idler) beam with the infrared pump laser has been shown to be consistent with multi-Watt output powers. This extends the OPO output at high output power and stability into the visible regime [6,7]. As an attractive feature of periodically poled crystals, it is possible to implement the OPO and SFG processes on the same PPLN crystal, greatly reducing the complexity and challenge of aligning the device for simultaneous phase-matching of two processes. This permits also some tunability by designing different poling periods in the same PPLN crystal. Bosenberg et al. demonstrated this combined OPO-SFG approach (with poling periods of fixed length) in a Nd:YAG pumped OPO, yielding output power in the regime beyond 1 Watt, at moderate amplitude noise in the regime of some percent during one minute, and temperature tunability of roughly 7 nm around a center wavelength of 629 nm . However, the output power dropped after some 10 hours of operation due to perturbations associated with the photorefractive effect in lithium niobate . In our fiber laser pumped OPO-SFG, we use magnesium-oxide-doped PPLN instead, which does not suffer from degradation on long time scales and hence provides stable output of the OPO-SFG. Moreover, the OPO of Bosenberg et al. was spectrally broad due to the use of a multi-frequency pump laser.
We note, that the frequency jitter from free running cw-OPOs is typically too large to directly apply these systems in high-precision experiments, e.g. in high-resolution spectroscopy or quantum optics. Thus, additional frequency stabilization is required. Recent attempts already demonstrated promising frequency stabilization of cw-OPOs , as well as cw-OPOs with intra-cavity second harmonic generation to values below 1 kHz . We follow similar approaches to reduce the OPO linewidth. In our setup we apply a fiber laser to pump the OPO system. The pump wave drives a signal wave and an idler wave in the mid-infrared spectral region. The signal wave is afterwards mixed with the pump wave to provide visible output. Mixing the pump wave with the signal wave transfers also the frequency noise of the pump to the SFG output. However, fiber pump lasers provide typically very low frequency noise, e.g. well below 100 kHz. Thus we add only little frequency noise, while maintaining the possibility to tune the SFG output via the pump wavelength. We extend the system by a Pound-Drever-Hall (PDH) frequency stabilization  to obtain visible radiation in the range between 605 nm and 616 nm with frequency jitter in the range of 100 kHz. The setup yields output powers well beyond 1 W, maintaining single-longitudinal mode tunability and long-term stability for many hours.
The OPO-SFG system is meant for applications at a wavelength of slightly below 606 nm in solid-state quantum optics, i.e. to coherently drive an optical memory in a Praseodymium-doped yttrium-orthosilicate (Y2SiO5) host crystal. The latter is a standard medium for optical data storage and implementations of quantum memories. Recently we applied Pr:Y2SiO5 for stopped light and image storage by electromagnetically-induced transparency (EIT), reaching ultra-long storage times up to one minute . So far, the required narrow-band radiation at 606 nm with sufficient output power to drive Pr:Y2SiO5 is only available from dye lasers or frequency generation of two phase-locked lasers. The OPO-SFG approach enables an alternative, all-solid-state setup of a laser-driven optical memory.
2. Experimental setup
The experimental setup of the OPO-SFG is based on a commercially available cw-OPO system, which originally provides mid-infrared signal radiation in the range of 1.4 - 2 µm (Lockheed Martin Aculight Argos Model 2400 SF-15). We modified the cw-OPO module towards OPO-SFG operation and frequency stabilization. The system is schematically depicted in Fig. 1 (see the upper dashed box). A fiber laser system, involving a seed fiber laser (NKT Photonics Koheras AdjustiK Y10) and a fiber amplifier (IPG YAR-15K-LP-SF) serves as a pump source and provides radiation at λP = 1064 nm with output powers of up to 16 W and a frequency jitter below 100 kHz. A piezo electric transducer (PZT) in the seed laser allows for continuous tuning of the fiber laser wavelength in a range of 100 GHz around λP = 1064 nm, as well as for frequency stabilization. The pump light is guided by an optical fiber to the OPO-SFG module, including a periodically poled, magnesium-oxide-doped lithium niobate crystal (MgO:PPLN, dimensions 50 × 12 × 0.5 mm) as the nonlinear medium.
The modifications from the commercially available OPO with mid-infrared output consisted solely of replacement of the OPO crystal with a crystal poled for both OPO and SFG processes, as well as replacing two dichroic mirrors behind the system to separate the different wavelengths. The new MgO:PPLN crystal contains two stages. The OPO process takes place in the first stage (length 42.5 mm), with “fan-out” poling structure, i.e. the poling period varies linearly from 27.2 µm to 28.9 µm perpendicular to the beam propagation direction. This enables coarse tuning the signal wavelength (and the idler wavelength λI correspondingly) of the OPO at e.g. a crystal temperature of T ≈40°C from λS = 1386 to 1445 nm by shifting the position of the MgO:PPLN crystal in the resonator (as indicated in Fig. 1). An oven permits variation of the crystal temperature up to 70° C as an additional parameter for slow frequency tuning of the OPO-SFG system. The SFG process takes place in the second stage of the MgO:PPLN crystal (length 7 mm). The stage contains three sections of constant poling periods (10.18 µm, 10.43 µm, and 10.69 µm), separated perpendicular to the beam propagation direction. The poling periods were calculated with the SNLO software (AS Photonics), based on the Sellmeier equations. We choose periods to obtain a large tuning range around a desired wavelength of λSFG = 606 nm(generated in the second track, i.e. with a poling period of 10.43 µm at a calculated temperature of 52°C by sum frequency mixing of the pump wave at λP = 1064 nm with a signal wave at λS = 1407 nm). We note that within one track of the SFG stage on the crystal, various output wavelengths can be generated by changing the crystal position and hence effectively tuning the generated signal wavelength in the preceding OPO stage with continuously changing poling periods. In addition, adjusting the crystal temperature permits changes of the SFG output wavelength by simultaneous phase-matching of both OPO and SFG processes. Thus, there is a broader range of possible output wavelength from each track. All three tracks were designed to cover in total an output wavelength range between 605 and 616 nm with a temperature range of 40°C to 140°C. We measure the SFG output wavelength with a wavelength meter (Angstrom High finesse WS6, not shown in Fig. 1).
We note, that previous work on related subjects (see Ref [6,7].) indicated, that crystals with reversed order of SFG and OPO stage should yield enhanced stability. While this has been tested in more recent versions of the OPO-SFG system, we did not yet perform systematic measurements to compare the performance of OPO-SFG versus SFG-OPO crystals in our system. This will be subject of future work.
The OPO-SFG crystal is embedded in a bow-tie optical ring cavity, which is singly resonant for the signal wave. The cavity consist of two curved mirrors (radii of curvature r = 100 mm) and two planar mirrors. The configuration of the cavity is essentially the same as in Ref . The mirrors are highly reflective with RS ~99.9% for the signal and RI,P < 5% for the idler and pump. The pump beam is focused to a 1/e2 diameter of 130 µm in the center of the crystal. For further details on the cavity see . We note that the mirrors are not specifically anti-reflection coated for visible wavelengths and therefore cause power losses of 20-30% for the generated SFG radiation. An additional 250 µm thick, fused silica low finesse etalon with a free spectral range of 400 GHz is placed between the two planar mirrors. Tuning of the etalon by changing the tilt angle (in a typical range of 8°) permits to roughly frequency tune the signal wave and hence also the visible SFG output. Fine tuning of signal and SFG output in the range of some 100 MHz is accomplished by a PZT (Physik Instrumente P-010.00H) on one of the cavity mirrors. For robust experimental handling, i.e. without changes on the cavity, it is also possible to tune the SFG output frequency by a PZT in the fiber seed laser, i.e. by changing the pump wavelength. A computer and homemade software serves to simultaneously control the relevant parameters of the OPO laser system, e.g. pump power, etalon angle and crystal temperature. This enables quick automatic turn-on and alignment of the laser system for an arbitrary wavelength in the tuning range.
We extend the OPO-SFG by a frequency stabilization for the visible SFG radiation, based on the Pound-Drever-Hall (PDH) method  (see lower dashed box in Fig. 1). The PDH stabilization involves an external Fabry-Perot resonator (dark grey box in the lower dashed box in Fig. 1) with a finesse Ƒ ≈3000, a length of l = 145 mm and curved mirrors (radii of curvature r = 1000 mm). To lock the laser system, a fraction of the visible SFG output is guided through an electro-optical modulator (Thorlabs EO-PM-NR), driven with a modulation frequency of f = 6.25 MHz. The phase-modulated beam is then spatially mode matched by a single mode fiber and a focusing lens to the TEM00 mode of the PDH resonator. A fraction of the beam enters the PDH resonator, building up a longitudinal mode, whereas the major part is promptly reflected and superimposed with the leakage beam of the PDH resonator (as indicated by the arrows in Fig. 1). The combination of a λ/4 plate and a polarizing beam splitter cube serves to spatially separate the incoming and the back-propagating beam. The latter is detected with a fast photodiode (Motorola MRD 510). The electronic signal is amplified (Femto DHPVA-100) and sent to a locking electronic unit (Toptica Digilock 110). The unit demodulates the signal and after appropriate filtering sends it to high voltage amplifiers (Piezomechanik SVR 200/3), which drive the PZTs in the OPO cavity and the seed laser.
3. Experimental results and discussion
3.1 Free running OPO-SFG system
We now investigate the performance of the OPO-SFG system without external PDH stabilization. At a pump power of Pp = 11.3 W the system provides stable SFG output power of PSFG = 1.3 W at a wavelength of λSFG = 606 nm, i.e. the relevant optical transition wavelength in Pr:Y2SiO5. To the best of our knowledge, this exhibits the highest output power for experiments at this specific wavelength from an all-solid-state laser system. We measured a rather high threshold pump power of Pthr ≈8 W to yield visible output. This rather high threshold is due to absorption of the idler radiation at λI = 4365 nm (absorption coefficient 0.3/cm) and depletion of the signal wave in the SFG stage of the crystal.
We now consider coarse tuning in a broader wavelength range. This works by variation of the MgO:PPLN crystal position, crystal temperature and etalon tilt angle. Figure 2(a) shows the possible SFG output wavelength, when the MgO:PPLN crystal position is varied, and the etalon angle as well as the crystal temperature are adjusted manually for each crystal position. The pump power was kept constant at Pp = 15 W. The total range covers an interval of Δλ > 10 nm from λ SFG = 605 nm to λ SFG = 616 nm. The SFG output power in this interval is always well above 1 W but varies with the exact SFG output wavelength. The power variation is primarily due to absorption of the PPLN crystal at signal and idler wavelengths. In particular, the idler absorption increases from 3% per cm to 30% per cm in the wavelength range from 3900 nm to 4400 nm and thereby causes thermal dephasing and lensing, reducing the OPO and SFG efficiency. The intrinsic single pass signal absorption of the PPLN crystal is indeed very low. However, there is also significant localized absorption (particularly at λS ≈1450 nm) of the signal beam due to OH content in the crystal which reduces the intra-cavity circulating power and thus the SFG efficiency (our own measurements of these features have shown that such OH absorption can be greatly reduced by high temperature annealing of the nonlinear crystal).
Moreover, the lack of anti-reflection coatings for the SFG output on the resonator mirrors causes additional variations of the output power with wavelength. Despite these effects, we nevertheless observe a broad spectral range of high power SFG output. We note, that also the quality of the Gaussian-shaped SFG output beam is very good, typically yielding M2 values of 1.05 in X-direction and 1.10 in Y-direction (measured with a beam profiler at output wavelength λ SFG = 611 nm).
Finally we remark, that for longer SFG output wavelengths the idler absorption in the crystal decreases and the OPO and SFG efficiency would be more uniform. This allows application of the OPO-SFG concept (with a different appropriately poled crystal) to tune an OPO-SFG system with signal wavelengths between 1460 nm and 2060 nm, yielding SFG output up to λ SFG ≈700 nm. Likewise the same concept can be applied in resonating and summing with the idler wavelength, providing SFG output in the 700 nm to 800 nm region.
We now fix the position and temperature of the MgO:PPLN crystal, as well as the etalon angle, while varying the pump laser frequency via the PZT in the fiber seed laser. This enables fast and fine frequency tuning of the OPO-SFG system, i.e. precise choice of the SFG output wavelength. Figure 2(b) shows the variation of the visible SFG output wavelength (red, solid line), when the PZT in the pump seed laser is driven with a triangular modulation voltage (black, dashed line). Thereby the signal frequency in the OPO cavity stays the same, while the SFG output frequency changes with the pump frequency. Thus, the SFG output wavelength exactly follows the modulation. The setup provides more than 20 GHz mode hop free tuning range around λ SFG = 606 nm. This is already sufficient to cover, e.g. the full inhomogeneously broadened optical linewidth of Γinh = 7 GHz in Pr:Y2SiO5.
To monitor the stability of the SFG output power and SFG wavelength, we set the wavelength of the free-running system (i.e. without any external frequency stabilization) closely below 606 nm and performed a long-term measurement of these quantities for many hours (see Fig. 3(b)). The free running OPO-SFG system provides quite stable output for more than 10 hours without mode hops or sudden jumps of wavelength and output power. During this long period, the drift in the output power is in the range of ΔP/P = 8%. On short time scales of some seconds, power fluctuations are much smaller, i.e. in the range of ΔP/P < 1%. The variation of the central wavelength is below Δλ = 0.5 pm, corresponding to frequency drifts of Δν = 400 MHz. The interrupt at 1.5 h is due to a parallel measurement to prove single-longitudinal mode operation (see Fig. 3(a)), which slightly perturbed the long-term measurement.
After more than 10 hours a first longitudinal mode hop occurs and the system starts to run in longitudinal multi-mode. An increase of average output power precedes this mode hop. This probably leads to additional heating of the crystal after 10 hours and hence a small variation of the quasi-phase-matching condition, which triggers multi-mode operation. However, by adjusting the pump power, crystal position or temperature, multi-mode operation probably could have been suppressed even after 10 hours.
Besides the drifts of wavelength and output power on long time scales, the data show oscillations on timescales of roughly half an hour. These are due to feedback control oscillations in the temperature of the air-conditioned laboratory. This affects the crystal temperature and hence wavelength and output power. Better stabilization of the OPO temperature (or external frequency and intensity stabilization) would suppress the oscillations and yield an even better stability of output power and wavelength on long timescales. Despite these oscillations, which are no intrinsic problems of the OPO-SFG, high output power and frequency stability with already quite low jitter are maintained over several hours of operation, even in the free-running system and under imperfect laboratory conditions.
3.2 OPO-SFG system with PDH frequency stabilization
To improve the frequency stability of the OPO-SFG system, we apply now the PDH frequency stabilization. To precisely measure the expected narrow OPO-SFG linewidth, we performed a spectral hole-burning (SHB) experiment in Pr:Y2SiO5. In the following we give a very brief overview of the spectroscopic part of the experiment. For a detailed description of SHB in rare-earth ion doped crystals such as Pr:Y2SiO5, we refer the reader to Refs [14,15]. We apply a frequency chirped pump pulse to prepare a broad spectral hole in the absorption spectrum (i.e. a “spectral pit”) of the inhomogeneously broadened manifold of Praseodymium ions. A second repump pulse transfers atomic population back into the spectral pit, and thus generates spectrally narrow absorption lines (see Fig. 4(a)). A third pulse is used to probe (i.e. to record) the absorption. Acousto-optic modulators generate all pulses from the cw output of the OPO-SFG system. The spectral linewidth of the absorption lines is determined by the homogeneous linewidth, residual saturation broadening, Fourier bandwidth and laser linewidth of the repump and probe pulses. Residual saturation broadening can be calculated from the measured laser intensity and the known transition moments in Praseodymium ions, or very easily deduced from other spectroscopic experiments on Pr:Y2SiO5 in our laboratory. The Fourier bandwidth is determined by the pulse shape and pulse duration. Thus, the width of the absorption lines in the spectral pit directly reveals information on the laser linewidth of repump and probe laser pulses generated by the OPO-SFG system.
Figure 4(a) shows a spectral pit with absorption lines in Pr:Y2SiO5, generated and probed by laser pulses from the frequency stabilized OPO-SFG system. The absorption lines appear spectrally very sharp. We note, that without frequency stabilization, preparation and probing of the spectral pit was hardly possible at all. From a single absorption line (see Fig. 4(b)) we deduce an absorption linewidth of Δν = 183 kHz (FWHM). To estimate the laser linewidth ΔνSFG from the spectroscopic data, we assume the measured absorption line to be a convolution of the following spectral lineshapes: (a) Two subsequent saturation broadened transitions, driven by the re-pump and probe process, with saturation broadening of Δν = 23 kHz (re-pump) and Δν = 16 kHz (probe), calculated from spectroscopic parameters in Pr:Y2SiO5 . Our many previous experiments in Pr:Y2SiO5 with other lasers confirm the calculated saturation broadenings. (b) Laser linewidths, i.e. the spectral frequency noise distribution in the re-pump and probe pulses. We assumed the frequency noise to be the same in both pulses, as they both come from the same OPO-SFG system. (c) Fourier bandwidth of the long pulses below 1 kHz, i.e. essentially negligible here. Convolution of the four contributions leads to a Voigt profile, which we compare to the measured absorption line to deduce an upper limit for the laser linewidth of ΔνSFG ≈114 kHz (FWHM).
To confirm this value by an alternative method, we also measured the laser linewidth directly by monitoring the time-resolved side-of-fringe transmission through a low finesse (Ƒ ≈100) Fabry-Perot resonator (Sirah Eagle Eye). In a first step, the resonators length is scanned with a PZT (attached to one of the resonators mirrors) to record transmission in one full free spectral range. This serves to determine the maximally transmitted power. The resonator length is then again varied, until half of the amplitude is detected after the resonator, i.e. we monitor one side of the transmission fringe. In the next step, the resonator length is kept fixed and the device monitors the time-resolved, transmitted power through the resonator (in our case for a period of 100 ms). This power changes, when the frequency of the laser varies within the side of the transmission peak of the resonator. Consequently, the device monitors frequency jitter of a laser as jitter in the power, transmitted through the resonator.
Figure 5 shows the Fourier transform of the raw transmission data (i.e. the frequency noise power spectral density), for the free running as well as for the frequency stabilized OPO-SFG system. We note, that the noise power spectral density does not show specific nor high frequency noise for the OPO-SFG system. Thus, already the free running OPO-SFG permits stable operation (specified by a frequency jitter below 1 MHz). The PDH frequency stabilization strongly reduces the jitter, as the noise power spectral density shows. Processing the raw data in the commercial device yields a laser linewidth of ΔνSFG ≈60 ± 20 kHz (FWHM), measured on a timescale of 100 ms. This is even below the upper limit, determined from the absorption experiments in Pr:Y2SiO5. Further improvements in the PDH stabilization, e.g. by taking the PZT resonance frequencies into account , should even permit better frequency stability. Nevertheless, already the laser linewidth obtained with the PDH stabilization as is, permits very narrow-band operation of the OPO-SFG system, i.e. with fully sufficient frequency resolution, e.g. for applications in quantum optics.
We report on the implementation of an all solid-state, singly resonant optical parametric oscillator (OPO), pumped by a fiber laser with low frequency noise, combined with intra-cavity sum frequency generation (SFG) to obtain visible radiation, and extended by a simple Pound-Drever-Hall (PDH) frequency stabilization unit. The still rather compact system is easy to align, to operate and to tune in frequency, also under robust and realistic experimental conditions. The OPO and SFG processes use a single, custom-made, periodically poled lithium niobate (PPLN) crystal, divided into sections with appropriately designed poling periods. The OPO-SFG system provides single-longitudinal mode radiation, at high output power well above 1 Watt, in a spectral range from 605 nm to 616 nm. With alternative, appropriately poled PPLN crystals and resonator optics (e.g. for a signal- or idler-resonant cavity), the OPO-SFG system may easily modified to generate similar output at any wavelength between 600 nm and 800 nm. Our particular system is meant for experiments in solid state quantum optics, i.e. to implement optical memories in Pr:Y2SiO5 crystals, driven on an optical transition close to 606 nm. The PDH stabilization reduces the laser linewidth of the OPO-SFG system below roughly 100 kHz. We determine upper limits for the laser linewidth and prove narrow-band operation of the OPO-SFG system by measurements of high-resolution spectral hole burning in Pr:Y2SiO5. The system will be attractive for applications in high-precision laser spectroscopy, atomic and molecular physics, or quantum optics.
The authors acknowledge valuable comments and discussions by H. Luckmann (University of Düsseldorf), T. Feldker (University of Mainz) and T. Channeliere (Laboratoire Aimé Cotton, Orsay). The research leading to these results has received funding from the Deutsche Forschungsgemeinschaft, and the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme FP7/2007-2013/ under REA Grant No. 287252.
References and links
1. D. D. Arslanov, M. Spunei, J. Mandon, S. M. Cristescu, S. T. Persijn, and F. J. M. Harren, “Continuous-wave optical parametric oscillator based infrared spectroscopy for sensitive molecular gas sensing,” Laser Photon. Rev. 7(2), 188–206 (2013). [CrossRef]
2. J. A. Armstrong, N. Bloembergen, J. Ducuing, and S. Pershan, “Interactions between Light Waves in a Nonlinear Dielectric,” Phys. Rev. 127(6), 1918–1939 (1962). [CrossRef]
3. 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 LiNbO,” J. Opt. Soc. Am. B 12(11), 2102–2116 (1995). [CrossRef]
4. J.-M. Melkonian, T.-H. My, F. Bretenaker, and C. Drag, “High spectral purity and tunable operation of a continuous singly resonant optical parametric oscillator emitting in the red,” Opt. Lett. 32(5), 518–520 (2007). [CrossRef] [PubMed]
5. O. Mhibik, D. Pabœuf, C. Drag, and F. Bretenaker, “Sub-kHz-level relative stabilization of an intracavity doubled continuous wave optical parametric oscillator using Pound-Drever-Hall scheme,” Opt. Express 19(19), 18049–18057 (2011). [CrossRef] [PubMed]
7. G. T. Moore and K. Koch, “Optical parametric oscillation with intracavity sum-frequency generation,” IEEE J. Quantum Electron. 29(3), 961–969 (1993). [CrossRef]
8. F. S. Chen, “Optically Induced Change of Refractive Indices in LiNbO3 and LiTaO3,” J. Appl. Phys. 40(8), 3389–3396 (1969). [CrossRef]
9. R. Al-Tahtamouni, K. Bencheikh, R. Storz, K. Schneider, M. Lang, J. Mlynek, and S. Schiller, “Long-term stable operation and absolute frequency stabilization of a doubly resonant parametric oscillator,” Appl. Phys. B 66(6), 733–739 (1998). [CrossRef]
10. R. W. P. Drever, J. L. Hall, F. V. Kowalski, J. Hough, G. M. Ford, A. J. Munley, and H. Ward, “Laser phase and Frequency Stabilization Using an Optical Resonator,” Appl. Phys. B 31(2), 97–105 (1983). [CrossRef]
11. G. Heinze, C. Hubrich, and T. Halfmann, “Stopped Light and Image Storage by Electromagnetically Induced Transparency up to the Regime of One Minute,” Phys. Rev. Lett. 111(3), 033601 (2013). [CrossRef] [PubMed]
12. W. R. Bosenberg, A. Drobshoff, J. I. Alexander, L. E. Myers, and R. L. Byer, “93% pump depletion, 3.5-W continuous-wave, singly resonant optical parametric oscillator,” Opt. Lett. 21(17), 1336–1338 (1996). [CrossRef] [PubMed]
14. K. Holliday, M. Croci, E. Vauthey, and U. P. Wild, “Spectral hole burning and holography in an YSiO:Pr crystal,” Phys. Rev. B 47(22), 14741–14752 (1993). [CrossRef]
15. M. Nilsson, L. Rippe, S. Kröll, R. Klieber, and D. Suter, “Hole-burning techniques for isolation and study of individual hyperfine transitions in inhomogeneously broadened solids demonstrated in Pr3+:Y2SiO5,” Phys. Rev. B 70(21), 214116 (2004). [CrossRef]