Open Access
Add to BibSonomyAbstract
We report a quasi-continuous-wave external cavity Raman laser based on potassium yttrium tungstate (KYW). Laser output efficiency and spectrum are severely affected by the presence of high gain Raman modes of low frequency (< 250 cm−1) that are characteristic of this crystal class. Output spectra contained frequency combs spaced by the low frequency modes but with the overall pump-to-Stokes conversion efficiency at least an order of magnitude lower than that typically obtained in other crystal Raman lasers. We elucidate the primary factors affecting laser performance by measuring the Raman gain coefficients of the low energy modes and numerically modeling the cascading dynamics. For a pump polarization aligned to the Ng crystallo-optic axis, the 87 cm−1 Raman mode has a gain coefficient of 9.2 cm/GW at 1064 nm and a dephasing time T2 = 9.6 ps, which are both notably higher than for the 765 cm−1 mode usually considered to be the prominent Raman mode of KYW. The implications for continuous-wave Raman laser design and the possible advantages for applications are discussed.
© 2016 Optical Society of America
1. Introduction
Solid state Raman lasers are convenient and efficient devices for extending the wavelength range of inversion based lasers [1]. External cavity Raman lasers (ECRLs) offer simplicity and flexibility over other configurations for designing the resonator in terms of thermal-lens management and mirror reflectivities by placing the Raman crystal in a separate cavity singly resonant to the Stokes wavelength [2]. However, achieving continuous-wave (CW) operation is challenging as high-intensity pumping, small mode sizes and low-loss resonators are required to achieve moderate lasing thresholds [3–5]. The first CW ECRL used barium nitrate and was limited to 5% conversion efficiency and 164 mW output power due to strong thermal lensing in the crystal [3]; whereas more recent work using diamond has enabled conversion efficiency exceeding 60% and output powers up to 380 W due to diamond’s excellent thermal properties and high Raman gain [6]. However, since the Raman shift in diamond is relatively large (1332 cm−1), there is still demand for crystalline Raman lasers with greater flexibility in terms of output wavelength, despite having less output power potential.
The class of double metal tungstates—of which potassium gadolinium tungstate (KGd(WO4)2, KGW) and potassium yttrium tungstate (KY(WO4)2, KYW) are prominent examples—offer the distinct advantage of two strong phonon modes around 765 cm−1 and 905 cm−1 that have similar Raman gain coefficients and can be accessed separately by changing the polarization state of the pump radiation with respect to the crystallographic axes of the crystal [8, 9]. As shown in Table 1, they also have thermal conductivity approximately three times that of barium nitrate [11], a moderate Raman gain coefficient, a high optical damage threshold and robust mechanical properties [12, 13]. The two aforementioned modes correspond to the symmetric stretching vibrations of the WOOW and WO molecular groups, respectively [14–16]. Although most reports on double metal tungstate Raman lasers utilize these two phonon modes, others have been observed. Kaminskii et al. [9] and Hanuza and Macalik [16] reported strong Raman modes of frequency 87 cm−1 and 225 cm−1 in the spontaneous Raman spectrum of KYW that have been attributed to a librational mode of WOn and a translation mode of Y3+ ions, respectively. Laser operation on a low frequency mode (84 cm−1) was observed in an intra-cavity KGW Raman laser which was designed for lasing on the prominent 768 cm−1 mode [17]. In this work, the output wavelengths resulted successive Stokes shifts from the 768 cm−1 and 84 cm−1 modes. Following [18], we refer to this process of cascaded stimulated Raman scattering (SRS) via two or more different phonons as cross-cascading. Cross-cascading has also been observed in single-pass excitation of other crystals. On the other hand, lasing on the 225 cm−1 phonon mode has not been observed to our knowledge. Similar cross-cascading processes have been utilized for multi-wavelength Raman lasers in other double metal tungstate devices [19]. Therefore double metal tungstate crystals offer a greater choice of Stokes wavelengths than diamond or barium nitrate, in addition to reduced susceptibility to thermal effects compared with barium nitrate.
Table 1. Comparison of KYW Properties with Other Raman Crystals Used in CW-ECRLs
In this paper, we demonstrate a quasi-CW ECRL employing KYW pumped at 1064 nm to generate laser radiation at 1158 nm and 1177 nm, corresponding to the first Stokes components of the two prominent high-energy phonon modes. However, we find that cross-cascading involving the 87 cm−1 and 225 cm−1 Raman modes plays a critical role in determining the output spectrum and laser efficiency. Polarization-dependent spontaneous Raman spectra of the crystal have been recorded to understand the manifestation of the observed phonon modes for different pumping conditions, and a numerical model was developed to analyze the role of cross-cascading and its impact on ECRLs. The 87 cm−1 mode has been found to have a gain coefficient, significantly higher than the two modes often regarded as the primary Raman modes.
2. Experimental setup
2.1 Raman laser setup
The pump source was a polarized Nd:YAG laser, multimode at 1064 nm, delivering 200 W of power in 250 µs pulses at 40 Hz repetition rate and having M2 less than 1.5 [20]. A half-wave plate was utilized to align the pump polarization along either the Ng or Nm crystallo-optic axis. The pump beam was focused into a 50 mm long KYW crystal which was placed in a near concentric 120 mm long cavity, as shown in Fig. 1. The end faces of the KYW crystal were coated with broadband AR coatings from 1000 to 1200 nm to reduce reflection losses.
Fig. 2 shows the reflectance spectra of the input coupler and output couplers which formed the linear Raman oscillator, obtained using a spectrophotometer (Cary 5000). The input coupler (IC) was highly transmitting (T = 96.8%) at 1064 nm and highly reflective (R > 99.9%) at the first Stokes wavelength (1158 nm) and at the observed cascaded Stokes wavelengths which are indicated by vertical dashed lines in the figure. The following experiments were carried out using two different output couplers OC1 and OC2, with reflectance of 99.89% and 98.6% at 1158 nm, respectively.
Fig. 2 Reflectance of the input coupler (IC) and the two output couplers (OC1 and OC2) used in the KYW Raman laser. The measurement uncertainty is 0.2% and 0.5% for measurements of OC1 and OC2, respectively. The vertical dashed lines indicate the reflectance at pump and different Stokes wavelengths observed experimentally (as shown in Fig. 3(b)).
The Stokes output power was measured using a thermal power sensor. The output spectrum was recorded using a spectrometer (Ocean Optics NIR 512), and the spatial beam profile was measured employing a CCD camera (WinCamD, DataRay Inc.).
2.2 Raman microscopy
A high-resolution (< 1 cm−1) Raman spectrometer (LabRAM HR Evolution, HORIBA Ltd.) incorporating a 532 nm laser was employed to record the spontaneous Raman scattering spectra. The probe beam was propagated along the crystallographic b-axis (the lasing direction) with its polarization vector varied using a half-wave plate.
3. Results
3.1 Laser performance
For OC1, the first Stokes at 1158 nm had a threshold of 18 W. Above 60 W of pump power, the Stokes output increases with a slope efficiency of 0.1%, reaching a maximum conversion efficiency of 0.04% (see Fig. 3(a)). The efficiency was fifty times lower than that obtained with a CW barium nitrate ECRL [3] and a thousand times lower than in a CW diamond ECRL [6]. There was no evidence of output saturation and the spatial beam profile of the Stokes output remained TEM00 across the investigated power range (see inset of Fig. 3(a)). Therefore, the cause of the inefficiency was deduced to be primarily non-thermal in nature. The time constant for establishing steady-state thermal gradients in KYW for a pump spot radius of 23 µm is approximately 500 μs [20, 21], thus the laser pulses are in a regime of thermal non-equilibrium and the temperature gradients are even smaller than expected for the stated output power. Measurements of the residual pump power show that the amount of pump depletion in the Raman crystal was negligible, confirming that a non-thermal mechanism was inhibiting power transfer in the Raman crystal.
Fig. 3 (a). Output Stokes peak power versus pump peak power when using OC1. The pump polarization direction was parallel to the Ng axis. The inset shows the Stokes beam profile at 92 W pump power. (b). Laser output spectrum at 50 W pump power, containing the pump line at 1064 nm and various Stokes components. The brackets indicate the Raman modes responsible for the various output wavelengths (765 cm−1, 905 cm−1 and 87 cm−1, respectively).
The Raman laser output spectrum shown in Fig. 3(b) reveals, in addition to the first Stokes lines at 1158 nm and 1177 nm corresponding to 765 cm−1 and 905 cm−1 shifts, several emission lines spaced by approximately 10 nm which is consistent with the 87 cm−1 phonon mode in KYW. At pump powers near the Raman laser threshold, only the first Stokes line at 1158 nm was attained. However, when the pump power was slightly increased (less than 1.2 times threshold), cross-cascaded Stokes generation involving the 87 cm−1 mode occurred, generating radiation at 1168 nm. With further increase in pump power to about 26 W, the Stokes line at 1177 nm corresponding to the 905 cm−1 Raman mode was also observed, followed by a Stokes shift from 1177 nm to 1188 nm via the 87 cm−1 Raman mode. This is followed by further cascaded Stokes shifts from 1168 nm to 1182 nm and from 1182 nm to 1194 nm via 87 cm−1 Raman mode.
For OC2, which showed higher transmission at the Stokes wavelengths, the laser exhibited a higher threshold pump power of 90 W, as expected. The Stokes power increased linearly with a slope efficiency of 4%, as shown in Fig. 4. The maximum conversion efficiency was 2% which is comparable to the aforementioned CW ECRL based on barium nitrate [3]. As for OC1, thermal effects in the Raman laser were negligible and the output spectrum exhibited a cross-cascaded spectral line due to the 87 cm−1 mode. At 110 W pump power, the output spectrum contained the primary 765 cm−1 Stokes mode and a single cross-cascaded line at 1168 nm (see inset of Fig. 4).
Fig. 4 Output Stokes peak power versus pump peak power when using OC2. The laser output spectrum in the inset was measured at 110 W pump power and the pump intensity is attenuated using a long pass filter. (Note that the shoulders on the short-wavelength side of the more intense lines in the inset is an artifact of the spectrometer.) The brackets indicate the Raman modes responsible for the Stokes wavelengths (765 cm−1 and 87 cm−1, respectively).
In a further experiment, we oriented the pump laser polarization along the Nm crystallo-optic axis using OC1, which suppressed lasing on the 765 cm−1 shift and enabled first Stokes lasing via the 905 cm−1 Raman mode. The threshold was 42 W, while cross-cascading to 1209 nm involving the 225 cm−1 mode was observed for pump powers above 50 W (see Fig. 5). For 90 W of input power, the first cross-cascaded line was stronger than the 905 cm−1 shifted line which we attribute in part to a higher transmission for the longer wavelengths for this output coupler. The threshold pump power for the second 225 cm−1 Stokes shift to 1241 nm was 90 W.
Fig. 5 Output spectrum at 90 W pump peak power for the pump polarization aligned to the Nm axis and when using OC1. The spectrum shows the pump wavelength at 1064 nm and three Stokes-shifted wavelengths generated from the 905 cm−1 and cross-cascaded 225 cm−1 phonon modes. The brackets indicate the Raman modes responsible for the first and cascaded Stokes wavelengths (905 cm−1 and 225 cm−1, respectively).
Similar to the observation for a Ng pump polarization, the higher output coupling provided by OC2 led to an increase in threshold (to 120 W). Cross-cascading by the 225 cm−1 mode was observed at a pump power of 130 W.
3.2 Raman spectrum of KYW
Although the 87 cm−1 and 225 cm−1 Raman modes have been previously identified, their gain coefficients have not been previously reported. In order to determine the gain coefficients for these shifts, we recorded high-resolution spontaneous Raman spectra for the two principal crystallo-optic directions.
For the pump polarization oriented parallel to Ng, as shown in Fig. 6(a), there are three main lines at 87, 765 and 905 cm−1. The peak value of the 87 cm−1 mode is substantially higher than for the 765 cm−1, indicating a higher stationary Raman gain coefficient. As is clearly apparent in the figure, the linewidth of the 87 cm−1 mode is much narrower than that of the 765 cm−1 mode, which suggests that the higher gain coefficient is in part attributable to a much longer dephasing time. Voigt profiles were fitted to the individual peaks of the instrument-broadened spontaneous Raman spectra after baseline correction in order to determine the Raman linewidths as well as the ratio of the gain coefficients. The results, summarized in Table 2, show that the stationary Raman gain coefficient of the 87 cm−1 mode is higher than that of the primary modes. It has a Raman linewidth of 1.1 cm−1, corresponding to a dephasing time T2 ≈9.6 ps which, for comparison, is slightly longer than the first-order mode in diamond (T2 ≈7.1 ps), but shorter than for the primary (Ag) mode in barium nitrate (T2 ≈26 ps) [7]. For the pump polarization aligned parallel to Nm, shown in Fig. 6(b), the 905 cm−1 line has the highest peak value with the 225 cm−1 line slightly lower. The linewidths of the primary Raman modes at 765 cm−1 and 905 cm−1 are in agreement with previously published values to within measurement uncertainty [22]. By referencing to a reported value of 3.6 cm/GW in the literature for the 905 cm−1 gain coefficient [9] and comparing the relative peak intensities of the respective Raman modes, we determined the gain coefficients at 1064 nm for the 87 cm−1 and 225 cm−1 to be 9.2 cm/GW and 2.5 cm/GW, respectively.
Fig. 6 Spontaneous Raman spectra of KYW for the pump polarization (E) parallel to (a) the Ng and (b) the Nm crystallo-optic axis. The insets to (a) and (b) show fitted line shapes (after baseline correction) of the 87 cm−1 and the 225 cm−1 modes, respectively.
Table 2. Linewidths and Relative Peak Intensities of the Raman Modes for Ng and Nm axes
4. Model analysis
We have observed that the presence of the low-energy Raman modes has a profound effect on the quasi-CW performance of a KYW ECRL. However, according to previous reports of KYW and KGW Raman lasers in other configurations (i.e. either pulse-pumped, intra-cavity, or both) [12, 23, 24], highly efficient operation (many tens of percent) can, in principle, be obtained without noticeable impact from low-energy modes. In order to elucidate the disparity between our observations and previous work, we have developed a numerical model to describe the power performance and spectral properties of the KYW ECRL. It is based on coupled rate equations describing the amplification and depletion of the pump and different Stokes fields via SRS [25, 26]. In this model, the pump and Stokes intensities are considered uniform as a function of position inside the cavity, and anti-Stokes generation, Raman four-wave mixing and dispersion effects are neglected. The generalized equations for the pump and first Stokes fields are:
where Ipump is the initial pump intensity, ga and gb are the gain coefficients of the primary Raman modes at 765 cm−1 and 905 cm−1, respectively, gc denotes the gain coefficient of the 87 cm−1 mode, at the pumping wavelength, and ηi is the quantum defect for the ith Stokes shift from the pump. K is the spontaneous Raman scattering factor, lR is the Raman crystal length, trt is the round-trip time of the cavity, Ls is the dissipative loss excluding the out-coupling loss, and R1p and R2p, and R1a,b and R2a,b are the input and output coupler reflectivities at the pump and Stokes wavelengths, respectively. The mirror reflectivities at each of the Stokes wavelengths were obtained from transmission measurements of Fig. 2. The cavity length was 120 mm and the pump spot sizes were 23 µm and 39 µm for OC1 and OC2, respectively. The model agreed well with the experimental observations when the round trip intracavity dissipative losses (scatter and absorption) were taken to be in the range 0.8% to 1.0%. Ip is the intracavity intensity for the pump, Ia,b and Iai,bi are the intracavity intensities for the first Stokes and cascaded Stokes components respectively. Equations for cascaded Stokes fields were also used in the model, where the pump field is replaced by the first or next-lower order Stokes field. These intracavity intensities were solved numerically using Mathematica, and the corresponding output powers were calculated and compared to the experimental data. The best agreement was achieved when the ratio of the gain coefficients for the 87 cm−1 and 765 cm−1 modes with respect to the 905 cm−1 mode was set to 2.5 and 1.25, respectively, which are within the uncertainty range of the gain coefficients derived from the spontaneous Raman spectra (compare Tables 2 and 3).Table 3. Raman Gain Coefficients Obtained by the Numerical Model (for Pump Polarization Parallel to Ng, Pump Wavelength: 1064 nm) by Comparison with Experiment
The model yields the intensities for each Stokes wavelength as a function of pump power. For a visual comparison with the experimental laser spectrum along Ng, we plotted the model results as Lorentzian lines at each Stokes wavelength with a linewidth of 0.7 nm. The modelled spectra are shown as dashed lines in Fig. 7(a) and (b) for OC1 and 2, respectively, along with the measured spectra (solid lines). Since we used an ECRL configuration which is non-resonant at wavelengths near the pump, first Stokes laser operation at 1074 nm generated directly from the pump via the 87 cm−1 mode does not reach threshold.
Fig. 7 Measured spectra (solid lines) and corresponding modelled spectra (dashed lines) for (a) OC1 at different pump powers and (b) OC2 at 110 W.
In all cases the model qualitatively reproduces what was observed experimentally: the onset of cascaded Stokes shifts at low thresholds accompanied by poor pump depletion and poor conversion efficiency. Using the coefficients in Table 3, the calculated laser threshold at 1158 nm for OC1 was 17.8 W in close agreement with the experimental value of 18 W. The model also shows that, with only a slight increase in pump power of about 2 W, the threshold for cross-cascaded Stokes at 1168 nm is achieved. This leads to clamping of the intracavity first Stokes intensity which subsequently stalls the transfer of pump power to the first Stokes. This is confirmed by a stepwise increase in residual pump power (not shown here). For a further rise in pump power by 5 W, the first Stokes related to the 905 cm−1 mode (at 1177 nm) is also seen, followed by a cross-cascaded Stokes line at 1187 nm owing to the interaction with the 87 cm−1 mode. The OC1 reflectivities for all the Stokes wavelengths are high and approximately equal (99.7 ± 0.1%), and thus promote the cross-cascading process.
In the case of OC2, the calculated threshold for lasing at 1158 nm (765 cm−1 mode) is 95 W, which is close to the measured threshold of 90 W. The first Stokes component corresponding to the 905 cm−1 mode (at 1177 nm) is obtained after an increase in pump power of approximately 30 W. These values agree well with those measured, confirming the accuracy of the model and the deduced Raman gain coefficients. It is evident from the model that the number of cascaded Stokes wavelengths is larger for OC1 compared to OC2 owing to the former’s higher reflectivity.
Although cross-cascading leads to clamping of a first Stokes line accompanied by stalling of the Stokes conversion efficiency, the model also shows that for pump powers higher than the threshold for a further cascade leads to a large recovery in the slope efficiency. This is because the cross-cascaded Stokes intensity becomes clamped which in turn unclamps the first Stokes intensity and power transfer from the pump. Indeed, using the model it was found generally that the slope efficiency changes from low to high values in an alternating fashion for odd and even number of cascades. The reason the laser efficiency remains poor overall is because the threshold for the odd-order cascade is very close to the previous one causing the slope efficiency to remain low for most of the power range. Although model derived thresholds are highly sensitive to losses, it reproduces well the higher efficiency observed for OC2 compared to OC1 and confirms that this is facilitated by the offsetting of cross-cascading to higher powers.
5. Discussion
The experimental results and the model show that the low frequency Raman modes have a major impact on the performance of the KYW laser, due to their high gain and the high output coupler reflectivity at wavelengths within 250 cm−1 of the first Stokes lines. The resultant low-threshold cross-cascaded Stokes shifting obstructs the efficient conversion of power from the pump to the first Stokes line, thus limiting the overall conversion efficiency. Therefore the impact of these modes must be taken into account when designing Raman lasers based on KYW or crystals with similar Raman spectral features. However, there have been numerous reports of efficient Raman lasers using double-metal-tungstate crystals such as KYW and KGW (which has a very similar structure and Raman spectrum to KYW) [12, 17, 19, 23, 24, 27–31], and only one that considers cross-cascading as a power-limiting factor [32].
One possible reason that these devices exhibited higher efficiency and were not apparently severely impacted by cross-cascading is that they operate with pump polarization along the Nm axis where the gain coefficient of the 225 cm−1 mode is lower than the 87 cm−1 and 905 cm−1 Raman modes [12, 19, 23, 24, 27–29]. In addition to the lower gain for this mode, the cross-cascaded wavelength is about 20 nm apart from the first Stokes so the output coupler reflectivity for the cascaded shift may be more differentiated from that of the first Stokes. Nevertheless, cross-cascading from the first Stokes mode (901 cm−1) via the 204 ± 3 cm−1 mode in KGW (equivalent to the 225 cm−1 mode in KYW) has been reported in a nanosecond-pulsed ECRL [30]. In that case they could not achieve slope efficiency similar to that obtained in other ECRLs and attributed it mainly to thermal effects in the crystal and the poor pump beam quality (cross-cascading was not discussed).
There are several papers which have reported cross-cascading in KGW lasers with pump polarization along Ng. In the case of a ns-pulse pumped KGW ECRL with a low-reflectivity output coupler [32], cross-cascading via the 84 cm−1 mode was observed at elevated powers, coinciding with a sudden decrease in slope efficiency. This may be evidence that cross-cascading is playing a similar role in limiting efficiency as observed in our work. In some CW intracavity and self-Raman lasers operating with low output coupling, cross-cascading on the low frequency Raman mode has been observed, although any adverse effects of cross-cascading on the laser efficiency was not discussed [17, 31]. Since the boundary conditions for power flow in intracavity Raman lasers are different to ECRLs, we cannot directly compare their results with our observations. Finally, there are other reports of efficient KGW and KYW lasers operating in the same orientation in which cross-cascading was not observed [12, 19, 23, 24]. This could be due to the specific loss properties of the resonators used (e.g. mirror coatings) or insufficient resolution in the characterization of the output spectrum.
Typically cross-cascading would be considered a parasitic process that should be avoided to improve laser efficiency. This can be achieved by tailoring the OC reflectivity at the first and cross-cascaded Stokes wavelengths in order to extend the range of pump power between the first and cascaded-Stokes thresholds. However, this would require sophisticated mirror coatings with significant changes in reflectivity (tens of percent) over a spectral range of only a few to a few tens of nanometers. This is likely to be costly and impractical. Introducing wavelength selective elements such as etalons or gratings into the cavity is an alternative approach for suppressing cross-cascading.
Even though the low frequency Raman mode at 87 cm−1 reduced the efficiency of this laser, it may be advantageous for other applications. This high gain mode could be used to simultaneously generate multiple closely spaced wavelengths which may be of use in spectroscopy or in terahertz radiation generation at integer multiples of 2.6 THz (87 cm−1) through difference frequency mixing of first and/or cross-cascaded Stokes waves. Furthermore, Raman beam conversion for the purposes of either brightness enhancement [33] through the Raman beam cleanup effect, or beam combination [34], can be enabled with much lower quantum defect than with conventional Raman modes in crystals. This may pave the way for an alternative approach to diamond in terms of high-power Raman lasers with possible advantages arising from the ease of growth of large tungstate crystals with good damage threshold and moderate thermal properties.
6. Conclusion
We have demonstrated an external cavity quasi-CW KYW Raman laser operating at multiple Stokes wavelengths. Cross-cascading involving the 765 cm−1 mode and the 87 cm−1 mode was observed when the pump polarization was aligned along the Ng axis, while a similar effect was present via 905 cm−1 and 225 cm−1 modes when the pump polarization was parallel to the Nm axis. Numerical modelling and experimental results confirmed that the early onset of cascaded Stokes components severely limits the pump depletion and thus, lowers the overall conversion to the output Stokes beam. The gain coefficients and dephasing times of the low frequency modes were calculated from high-resolution spontaneous Raman spectra. It is concluded that for a Raman crystal with one or more strong, low-frequency Raman modes, cross-cascading can be a major factor for consideration in Raman cavity design, particularly for CW systems that rely on highly-resonant cavities. The high gain coefficient of the 87 cm−1 Raman mode in KYW (9.2 cm/GW at 1064 nm) indicates potential for future applications as multi-wavelength laser sources of high beam quality and as a low-quantum-defect Raman material for brightness enhancement and/or beam combination.
Funding
Australian Research Council Discovery Grant (DP130103799); US Air Force Research Laboratory (FA2386-15-1-4075); German Research Foundation (DFG) (LU2018/1-1).
Acknowledgment
The authors thank David Adams for his assistance with the Raman spectrometer.
References and links
1. H. M. Pask, “The design and operation of solid-state Raman lasers,” Prog. Quantum Electron. 27(1), 3–56 (2003). [CrossRef]
2. V. G. Savitski, I. Friel, J. E. Hastie, M. D. Dawson, D. Burns, and A. J. Kemp, “Characterization of single-crystal synthetic diamond for multi-watt continuous-wave Raman lasers,” IEEE J. Quantum Electron. 48(3), 328–337 (2012). [CrossRef]
3. A. S. Grabtchikov, V. A. Lisinetskii, V. A. Orlovich, M. Schmitt, R. Maksimenka, and W. Kiefer, “Multimode pumped continuous-wave solid-state Raman laser,” Opt. Lett. 29(21), 2524–2526 (2004). [CrossRef] [PubMed]
4. O. Kitzler, A. McKay, D. J. Spence, and R. P. Mildren, “Modelling and optimization of continuous-wave external cavity Raman lasers,” Opt. Express 23(7), 8590–8602 (2015). [CrossRef] [PubMed]
5. O. Kitzler, A. McKay, and R. P. Mildren, “Continuous-wave wavelength conversion for high-power applications using an external cavity diamond Raman laser,” Opt. Lett. 37(14), 2790–2792 (2012). [CrossRef] [PubMed]
6. R. J. Williams, J. Nold, M. Strecker, O. Kitzler, A. McKay, T. Schreiber, and R. P. Mildren, “Efficient Raman frequency conversion of high‐power fiber lasers in diamond,” Laser Photonics Rev. 9(4), 405–411 (2015). [CrossRef]
7. R. P. Mildren and J. R. Rabeau, Optical Engineering of Diamond (John Wiley & Sons, 2013), Chap. 8.
8. Y. Romanyuk, “Liquid-phase epitaxy of doped KY(WO4)2 layers for waveguide lasers,” Thesis, Swiss Federal Institute of Technology in Lausanne (2005).
9. A. Kaminskii, A. Konstantinova, V. Orekhova, A. Butashin, R. Klevtsova, and A. Pavlyuk, “Optical and nonlinear laser properties of the χ (3)-active monoclinic α-KY(WO4)2 crystals,” Crystallogr. Rep. 46(4), 665–672 (2001). [CrossRef]
10. A. Sabella, D. J. Spence, and R. P. Mildren, “Pump–probe measurements of the Raman gain coefficient in crystals using multi-longitudinal-mode beams,” IEEE J. Quantum Electron. 51(12), 1–8 (2015). [CrossRef]
11. I. V. Mochalov, “Laser and nonlinear properties of the potassium gadolinium tungstate laser crystal KGd(WO4)2: Nd3+-(KGW: Nd),” Opt. Eng. 36(6), 1660–1669 (1997). [CrossRef]
12. R. Mildren, M. Convery, H. Pask, J. Piper, and T. McKay, “Efficient, all-solid-state, Raman laser in the yellow, orange and red,” Opt. Express 12(5), 785–790 (2004). [CrossRef] [PubMed]
13. H. M. Pask, “Continuous-wave, all-solid-state, intracavity Raman laser,” Opt. Lett. 30(18), 2454–2456 (2005). [CrossRef] [PubMed]
14. L. Macalik, J. Hanuza, and A. Kaminskii, “Polarized Raman spectra of the oriented NaY(WO4)2 and KY(WO4)2 single crystals,” J. Mol. Struct. 555(1–3), 289–297 (2000). [CrossRef]
15. D. Kasprowicz, T. Runka, A. Majchrowski, and E. Michalski, “Low-temperature vibrational properties of KGd(WO4)2:(Er, Yb) single crystals studied by Raman spectroscopy,” J. Phys. Chem. Solids 70(9), 1242–1247 (2009). [CrossRef]
16. J. Hanuza and L. Macalik, “Polarized infra-red and Raman spectra of monoclinic α-KLn(WO4)2 single crystals (Ln= Sm—Lu, Y),” Spectrochim. Acta Am. 43(3), 361–373 (1987).
17. D. C. Parrotta, W. Lubeigt, A. J. Kemp, D. Burns, M. D. Dawson, and J. E. Hastie, “Continuous-wave Raman laser pumped within a semiconductor disk laser cavity,” Opt. Lett. 36(7), 1083–1085 (2011). [CrossRef] [PubMed]
18. A. Kaminskii, O. Lux, H. Rhee, H. Eichler, H. Yoneda, A. Shirakawa, K. Ueda, R. Rückamp, L. Bohatý, and P. Becker, “Manifestations of nonlinear optical effects in a novel SRS-active crystal—natural topaz, Al2(F1−x(OH)x)2SiO4: many-phonon χ (3)-lasing, more than sesqui-octave Stokes and anti-Stokes multi-wavelength comb lasing, cascaded and cross-cascaded χ(3)↔ χ(3) Raman-induced interactions under single-and dual-wavelength picosecond collinear coherent pumping, THG and combined SRS-promoting phonon modes,” Laser Phys. Lett. 10(7), 073001 (2013). [CrossRef]
19. R. P. Mildren and J. A. Piper, “Increased wavelength options in the visible and ultraviolet for Raman lasers operating on dual Raman modes,” Opt. Express 16(5), 3261–3272 (2008). [CrossRef] [PubMed]
20. R. J. Williams, O. Kitzler, A. McKay, and R. P. Mildren, “Investigating diamond Raman lasers at the 100 W level using quasi-cw pumping,” Opt. Lett. 39(14), 4152–4155 (2014). [CrossRef] [PubMed]
21. A. McKay, O. Kitzler, and R. P. Mildren, “Thermal lens evolution and compensation in a high power KGW Raman laser,” Opt. Express 22(6), 6707–6718 (2014). [CrossRef] [PubMed]
22. B. Denker and E. Shklovsky, Handbook of solid-state lasers: Materials, systems and applications (Elsevier, 2013).
23. D. Chunaev, T. Basiev, V. Konushkin, A. Papashvili, and A. Y. Karasik, “Synchronously pumped intracavity YLF–Nd–KGW picosecond Raman lasers and LiF: F−2 amplifiers,” Laser Phys. Lett. 5(8), 589–592 (2008). [CrossRef]
24. J. Jakutis-Neto, J. Lin, N. U. Wetter, and H. Pask, “Continuous-wave watt-level Nd:YLF/KGW Raman laser operating at near-IR, yellow and lime-green wavelengths,” Opt. Express 20(9), 9841–9850 (2012). [CrossRef] [PubMed]
25. S. Ding, X. Zhang, Q. Wang, J. Chang, S. Wang, and Y. Liu, “Modeling of actively Q-switched intracavity Raman lasers,” IEEE J. Quantum Electron. 43(8), 722–729 (2007). [CrossRef]
26. S. Ding, X. Zhang, Q. Wang, F. Su, S. Li, S. Fan, S. Zhang, J. Chang, S. Wang, and Y. Liu, “Theoretical models for the extracavity Raman laser with crystalline Raman medium,” Appl. Phys. B 85(1), 89–95 (2006). [CrossRef]
27. A. A. Demidovich, A. S. Grabtchikov, V. A. Lisinetskii, V. N. Burakevich, V. A. Orlovich, and W. Kiefer, “Continuous-wave Raman generation in a diode-pumped Nd3+:KGd(WO4)2 laser,” Opt. Lett. 30(13), 1701–1703 (2005). [CrossRef] [PubMed]
28. V. Lisinetskii, A. Grabtchikov, A. Demidovich, V. Burakevich, V. Orlovich, and A. Titov, “Nd: KGW/KGW crystal: efficient medium for continuous-wave intracavity Raman generation,” Appl. Phys. B 88(4), 499–501 (2007). [CrossRef]
29. L. Batay, A. Kuzmin, A. Grabtchikov, V. Lisinetskii, V. Orlovich, A. Demidovich, A. Titov, V. Badikov, S. Sheina, V. Panyutin, M. Mond, and S. Kück, “Efficient diode-pumped passively Q-switched laser operation around 1.9 μm and self-frequency Raman conversion of Tm-doped KY(WO4)2,” Appl. Phys. Lett. 81(16), 2926–2928 (2002). [CrossRef]
30. A. McKay, O. Kitzler, and R. P. Mildren, “High power tungstate-crystal Raman laser operating in the strong thermal lensing regime,” Opt. Express 22(1), 707–715 (2014). [CrossRef] [PubMed]
31. C. Y. Tang, W. Z. Zhuang, K. W. Su, and Y. F. Chen, “Efficient continuous-wave self-Raman Nd: KGW laser with intracavity cascade emission based on shift of 89 cm−1,” IEEE J. Sel. Top. Quantum Electron. 21(1), 142–147 (2015). [CrossRef]
32. R. Chulkov, V. Markevich, V. Orlovich, and M. El-Desouki, “Steady-state Raman gain coefficients of potassium-gadolinium tungstate at the wavelength of 532 nm,” Opt. Mater. 50, 92–98 (2015). [CrossRef]
33. A. McKay, O. Kitzler, and R. P. Mildren, “Simultaneous brightness enhancement and wavelength conversion to the eye‐safe region in a high‐power diamond Raman laser,” Laser Photonics Rev. 8(3), L37–L41 (2014). [CrossRef]
34. A. McKay, D. J. Spence, D. W. Coutts, and R. P. Mildren, “Non-collinear beam combining of kilowatt beams in a diamond Raman amplifier,” in Advanced Solid State Lasers, OSA Technical Digest (online) (Optical Society of America, 2014), paper ATu5A.1.
