We report a pulsed Raman laser at 1193nm based on synthetic diamond crystals with a record output power of 24.5 W and a slope efficiency of 57%. We compared the performance of an anti-reflection coated crystal at normal incidence with a Brewster cut sample. Raman oscillation was achieved at both room temperature and under cryogenic operation at 77 K. Modeling of these experiments allowed us to confirm the value of Raman gain coefficient of diamond, which was found to be 13.5 ± 2.0 cm/GW for a pump wavelength of 1030 nm.
©2011 Optical Society of America
Raman lasers are attractive ways to generate new frequencies and to generate high brightness Stokes beams due to the associated beam clean-up . Among the various Raman-active materials, diamond has attracted early attention due to its large Raman gain and wide transparency range . Diamond’s large thermal conductivity, κ = 2000 W/m.K , and low thermo-optics coefficient, dn/dT = 7.9x10−6 , make it a material of choice for generating very high power Stokes beams with excellent beam quality. Continuous progress in the growth of low-birefringence, low-loss diamond crystals [5,6] has improved the Raman laser output performances reported in past years . High slope efficiencies, up to 84%, have been reported under pulsed operation , with maximum pulse energy of 0.6 mJ . Intra-cavity Raman lasers using synthetic diamond crystals have been demonstrated recently, with time-averaged Stokes powers of 375 mW in pulsed operation , and 200 mW in continuous-wave regime , mostly limited by parasitic losses. Despite the favorable slope efficiency reported in Chemical Vapor Deposition (CVD) diamond as compared with other Raman crystals, the maximum Stokes output power published so far is limited to 2.2W .
In this paper, we report on higher average power pulsed Raman lasers. The optimized thermal management of our experimental setup allowed us to generate a maximum Stokes output of 24.5 W at a repetition rate of 40 kHz, only limited by optical damage to the anti-reflection (AR) coated diamond crystal. Crystals with Brewster cut facets were also compared to the AR coated diamond crystals. The external cavity Raman laser was pumped by a Q-switched cryogenic Yb-doped YAG laser, delivering up to 340 W with diffraction-limited beam quality, which is described in section 2. The design of the Raman cavity and the output performances are detailed in section 3. One critical parameter for accurate modeling and optimization of the Raman laser is the gain, g R. Published data range from 33 cm/GW  to 75 cm/GW for a pump wavelength of 532 nm . Modeling of the Raman output performances allows us to derive the value of the gain coefficient for diamond, leading to g R (λ p = 1030nm) = 13.5 ± 2.0 cm/GW, in agreement with recent data, as described in section 4. Finally, we report on cryogenic operation of our Raman laser as a potential way for further power scaling. The improved thermal performances at lower temperature have been used in several laser materials in order to generate multi-kW beams with excellent brightness [15–17]. In the same way the thermal performance of diamond improves at lower temperature, κ = 9000 W/m.K  and dn/dT = 2x10−7  at 80K. From our experiments, the Raman gain is unchanged from 300 K to 80 K, which is, to our knowledge, the first measurement of a cryogenic diamond Raman laser.
Our results represent an order of magnitude improvement in output performance for a diamond Raman laser; the output power is more than double compared to the highest value reported in any crystal Raman laser .
2. Cryogenically cooled Yb:YAG pump laser
2.1 Experimental setup
The design of the pump laser, shown on the left side of Fig. 1 , was based on earlier work in our group. The Yb:YAG crystals, cavity mirrors and acousto-optic Q-switch were identical to the system reported in reference . The crystals were mounted in a vacuum enclosure to prevent condensation at cryogenic temperature. All surfaces were anti-reflection coated with high-damage threshold coatings (Ion-Beam-Sputtering, from Advanced Thin Films, Boulder CO), with typical residual reflection < 0.05%.
The cavity was end-pumped through 45° dichroic mirrors (M1 on Fig. 1). The pump beams at 940 nm were delivered from water-cooled 400-W fiber-coupled diodes (IPG, model DLM400, fibers core diameter 400μm, 0.22NA). These high-brightness diodes allowed us to maximize the overlap of the pump beams with the fundamental mode of the cavity, which was the main limiting factor in the performance of the earlier system. The double-sided end-pumping geometry requires high isolation of the pump modules from back-propagating power. This was achieved by the absorption of the pump beam through each YAG crystal, >97%, and by the rejection of the laser signal at 1030 nm by mirrors M1, >99.95%. The total length of the cavity was 133 cm which led stable operation of the fundamental mode over the target operating power.
2.2 Output performance
The cavity was first characterized under continuous-wave (cw) operation, with the acousto-optic Q-switch removed. The maximum output power at 1030 nm was 550 W, corresponding to an optical efficiency of 79% with respect to the pump power delivered to the crystals. The temperature of the crystals remained under 90 K at nominal operation, the output power did not show any sign of thermal roll-over and was only limited by the available pump power. The output beam was linearly polarized with an extinction ratio >15 dB over the entire range from threshold up to the maximum output power. The beam quality was close to the diffraction limit, with M 2<1.1 up to an output power of 450 W. At 500 W, stronger thermal lensing degraded the beam along one axis, leading to M 2 = 1.1x1.2. The excellent brightness of the laser is a direct consequence of the high thermal conductivity and low thermo-optic coefficients of YAG at 77K [15,16]. The present results are competitive with the best performance reported so far for a Q-switched Yb:YAG oscillator .
Q-switched operation was achieved with a repetition rate in the range of 40-100 kHz. The beam characteristics were unchanged compared to cw lasing: for a given pump power, the output power was only 1% lower. In the present work, we deliberately limited the output power to 340 W to keep sufficient margin with respect to optical damage, even though higher energy and lower repetition rate are achievable with the current setup. At 40 kHz and 340 W, the pulse duration was 75 ± 5 ns, the pulse energy and peak power were 8.5 mJ and 91 kW respectively. Use of a half-wave plate in conjunction with a polarizing beam splitter allow for attenuation of the pump beam power without changing the beam quality or pulse temporal shape. This performance and the diffraction-limited beam quality make it an ideal pump source for the Raman laser cavity. The system has been operated reliably for several hundred hours at nominal power without degradation and with minimal realignment.
3. Raman laser experiments
3.1 Description of the setup
The Raman laser cavity is shown on the right side of Fig. 1. For simplicity, the linear cavity was comprised of two identical mirrors, so that the Stokes beam exited from both ends. The cavity was 50 mm long, the mirrors had a radius of curvature of 200 mm. The mirror mounts used piezo-electric actuators so that the cavity could be aligned while inside the vacuum enclosure. The pump beam was focused by lens L1, the diameter at the location of the center of the crystal was 304 μm, measured with a CCD camera; the pump mode was well matched to the diameter of the fundamental mode of the cavity, 316 μm. Mirrors M6 had a reflectance R = 83% for the Stokes beam, and high transmission at 1030 nm, T = 98%. The mirrors also had low reflectance for the second Stokes beam (R = 6% at 1400 nm), so that only the first Stokes would oscillate even at high power.
The different beam sampling for diagnostics used dichroic mirrors (M4, T = 99.9% @1030 nm, R = 99.95% @1193 nm) and uncoated wedges M5. The pump and Stokes spectra were measured with a 0.007 nm resolution optical spectrum analyzer. Acquisition of the temporal profiles was achieved with Silicon and InGaAs photodiodes, with rise times of <1 ns and <5 ns respectively. The power of the different beams was measured with thermal power-meters. The overall accuracy of the measured threshold and slope efficiencies was estimated to be better than 10%.
The 8 mm-long CVD diamond crystal (Element6, Cambridge MA) was oriented for propagation along the <110> direction. Anti-reflection coatings were deposited on the 2x4 mm2 surfaces. The design of the coatings was optimized to maintain good adhesion under thermal cycling over the range [77 K-600 K]. The reflectance was 0.6%, 0.08% and 1.6% at 1030 nm, 1193 nm and 1400 nm respectively. The 2mm edge was parallel to the  axis, and the pump beam was polarized parallel to the 4mm edge, i.e. along .
One critical issue for power scaling is efficient thermal management of the gain medium. The single crystal diamond was soldered on a polycrystalline diamond heat-spreader (25.4x25.4x1 mm3). Silicon thermal sensors were epoxied on the top surface of the single crystal diamond and on the heat-spreader, as shown on Fig. 2 . This assembly was soldered on the metalized top surface of a BeO heat-sink. The heat-sink was attached on a copper cooler, where water or liquid nitrogen was circulated.
3.2 Optical quality of the diamond crystals
The optical quality of the single crystal diamond is an important parameter for the output performance of the Raman laser. We characterized the total losses of the samples, using a medium finesse cavity. Without crystal, the finesse was F = π/(1-R) = 6200 with R the reflectance of the mirrors. Inserting the CVD diamond in the cavity, the finesse was F’ = π/(1-R + L) = 605, leading to the total losses L = 0.5% at 1064 nm. Taking into account the reflectance of 0.08% per surface, the corresponding higher bound for the loss coefficient was 0.38%/cm, which includes absorption, scatter and losses due to wave-front distortion by the diamond crystal; it was very close to the best quality reported for CVD diamond .
The birefringence of the crystals was Δn = 5x10−6 measured by Element6 using birefringence microscopy (Metripol). We measured the spontaneous Raman gain in our crystals as a function of the polarization of the pump and Stokes beams. As expected for propagation along the  axis, a pump beam polarized along  generates a Stokes beam polarized along  . The extinction of the Stokes was better than 98%; this confirms that the birefringence of our sample was acceptable for laser applications since the depolarization was minimal along the 8 mm long crystal. We also verified that the same extinction was obtained throughout the 2x4 mm2 surface of the crystal.
3.3 Laser performances at room temperature
The output performances of the Raman laser are shown on Fig. 3 . The slope efficiency was 28.8% for one side, or 57.6% for the total Stokes output, which is lower than the one of Ref . However, the maximum output power measured with power-meter PM3 on Fig. 1 was 12.3 W at a repetition rate of 40kHz. This corresponds to a total Stokes power of 24.5 W, which is approximately an order of magnitude higher than the results reported for any diamond Raman laser so far . The stability of the Stokes average power was 2.2% at maximum output, as good as the stability of the pump laser, 2.5%. The maximum conversion efficiency is 13%, low compared to other reported systems, because we operated the laser only 1.5 times above threshold. The gain curve on Fig. 3 shows no sign of roll-over, so that higher pump power would increase the efficiency. The bandwidth of the Stokes beam varied between 0.042 and 0.048 nm over the entire range of output power; which is driven by the linewidth of the pump, 0.045 nm. The Stokes beam was linearly polarized along , parallel to the pump.
The temporal profiles on Fig. 4 show a very efficient depletion of the pump beam; this can also be seen on Fig. 3 where the transmitted pump power remains constant once the threshold of the Raman laser has been reached. The duration of the Stokes pulses was 29 ns, the maximum peak power was 19.2 kW.
Figure 3 shows no saturation and no sign of thermal roll-over. The heat load in the diamond crystal is given by. The first term is the quantum defect and the last two terms are the absorption of the incoming pump beam and of the intra-cavity Stokes power. Figure 5 shows the linear rise of the temperature measured by the two thermal sensors as a function of the heat load in the crystal. The temperature difference between the single crystal and the heat-spreader has a slope of 0.116 K/W. The value of this temperature rise can be calculated from the thermal resistance of the solder. A perfect indium bond would have a resistance of 3.1x10−6 W.m2/K at room temperature ; this would correspond to a slope of 0.097 K/W for our crystals. The good agreement with the measured data shows that efficient heat transfer is achieved through our assembly, and confirms the potential for further power scaling.
No second Stokes was detectable in any of our experiments. The Stokes output power was limited by optical damage to the anti-reflection coatings; the maximum output pulse energy was 0.6 mJ, corresponding to a fluence of 4.6 J/cm2 on the diamond faces. Coatings with an optimized design can withstand 20 J/cm2 for 100ns pulses; mirrors with lower reflectivity would also reduce the circulating Stokes intensity and allow further power scaling.
3.4 Brewster-cut crystal
Brewster cut crystals could be an interesting alternative to anti-reflection coated diamonds; it was used in some of the recent works [8,10,11]. One of our single crystals was cut for propagation along the <110> direction with incidence at Brewster angle, 67.4°. The optical length was 6.3 mm for that sample. The pump beam was polarized along . The linear cavity used 2 mirrors with R = 98.8% at 1193 nm, and T = 98% at 1030 nm. For a repetition rate of 50kHz, the threshold was 334W and the slope efficiency 58%. The maximum Stokes output power generated from this cavity was only 2.8 W, limited by the high intra-cavity Stokes power due to the high reflectance of the mirrors.
The same cavity was tested with the anti-reflection coated crystal, the corresponding threshold was 119 W. The larger threshold for the Brewster cut crystal is a result of the shorter crystal length, and the larger beam size due to oblique incidence. For the 50 mm-long cavity with mirrors radius of curvature of 500 mm, the waist diameter is 406μm for normal incidence, and 406 x 978 μm2 at Brewster angle. The expected ratio of the Raman laser threshold is which is in good agreement with the measured ratio of 2.81. This confirms that no significant depolarization losses occurred in our crystals, as could be expected from the low birefringence. However, the Brewster cavity could be configured using a 4-mirror geometry with cylindrical optics to achieve a circular beam at the focus by invoking astigmatism correction at the concave turning mirrors .
4. Raman gain coefficient
4.1 Raman laser model
The output performances of the Raman laser were modeled by numerical integration of the time-dependent Raman laser equations , with the addition of a third equation for the second Stokes field. The spatial overlap of the pump beam with the fundamental mode of the cavity is calculated from Ref . The Stokes field was initialized to one photon per mode, and numerical integration was achieved with a 4th order Runge-Kutta solver (Matlab ode45). The losses in the diamond crystal were treated as bulk losses. The input parameters of the model included the duration of the pump pulses, the pump beam diameter, the reflectivities of the mirrors at all wavelengths. All values were available from our measurements, so the only adjustable parameter was the Raman gain coefficient. The model was used to fit the measured output performances of the Raman laser; Fig. 3 shows a good agreement for both the transmitted pump power and the generated Stokes. Figure 6 shows the calculated temporal profiles, averaged over 3 ns in order to match the resolution of the photodiodes. The model is in good agreement with the data in Fig. 4; the duration and build-up time of the Stokes pulses are well calculated. The fast oscillations are due to the beating of the longitudinal modes. The depletion of the transmitted pump pulses is also well modeled.
4.2 Raman gain coefficient of CVD diamond
The above fit gives an estimate of the Raman gain coefficient of g R = 15 cm/GW. A second data set was measured with a different cavity in order to confirm the gain coefficient with different parameters. The mirrors M6 had higher reflectivity, R = 98.8% at 1193 nm. Their radius of curvature was 500 mm, corresponding to a beam waist diameter 406 μm. The pump beam was focused to a waist of 423x368 μm, the pulse duration was 85 ns (FWHM) and the repetition rate was 50 kHz. The measured threshold was 119 W; using the above model a good fit of the output performances was achieved with g R = 12 cm/GW. The average value of the gain coefficient is then g R = 13.5 cm/GW. The associated uncertainty can be estimated from the threshold of the Raman laser, since P th∝(T + L)/g R; the relative error for the transmission of the output coupler is ΔT/T = 6%, the error on the losses is ΔL/L = 25% and the threshold is measured with an accuracy of ΔP th/P th = 5% (from the calibration of the meter and the repeatability between several experimental runs). The overall uncertainty for our measurement of the Raman gain coefficient is 15%, so that g R = 13.5 ± 2.0 cm/GW.
There is a significant variation among the published values of the Raman gain of diamond, due to variations in the quality of the samples as well as differences in the measurement conditions. Spontaneous Raman scattering was used for relative comparison of different crystals ; normalizing to CaCO3, Ba(NO3)2 or KGd(WO4)2 leads to g R = 53 ± 20cm/GW for λ p = 0.532μm. The dispersion of the Raman gain coefficient can be deduced from Miller’s delta . Using the Sellmeier equations of diamond , the above value corresponds to g R = 20.5 ± 7.7cm/GW for λ p = 1.03μm. Reference  reports g R>8cm/GW for λ p = 1.06μm using stimulated Raman scattering. The more recent reference  lists g R = 12cm/GW for λ p = 1.06μm. As a conclusion, our measurement is at the lower end of the published values of g R, compatible with all reported uncertainties.
5. Cryogenic operation
As mentioned above, the thermal performances of diamond improve at low temperature, which could allow one to generate high power Stokes beam with high brightness. However no high power Raman laser operating at low temperature has been reported so far. We used the above Raman laser cavity described in section 3.1 in order to study the output performances with cryogenic cooling. In the present experiment, the pump beam diameter was 323 μm, the repetition rate was 40 kHz and the pulse duration 85 ns. The threshold was 150 W and the maximum output power was 20.6 W. The data of Fig. 7 have been fitted with the model of section 4 in order to derive the Raman gain coefficient; the value of g R = 15 cm/GW is similar to the one at room temperature derived from Fig. 3. No direct measurement of the Raman gain at 80K had been published earlier; however, the Raman gain bandwidth was found to have no measurable variation between 80K and 300K , which is in good agreement with the present results.
The output Stokes power in Fig. 7 shows thermal saturation at high power. The temperature of the diamond crystal varied between 86.5 K and 90.5 K over the range of Stokes power; it increased linearly at low power, and it had a super-linear behavior at high power. The increased thermal lensing also resulted in larger instability of the Stokes average power, 8%, almost 4 times more than at room temperature. The increased sensitivity to thermal effects is explained by the thermal properties of diamond. For pulsed operation, the figure of merit in Table 1 is actually lower at cryogenic temperature due to the lower heat capacity. Because the time constant for heat diffusion is longer than the Stokes pulses the temperature rise is mostly localized inside the interacting beams. A given heat load results in a stronger thermal lensing at cryogenic temperature; reduced overlap between pump and Stokes beams reduces the efficiency. This is no longer true in cw operation, where the figure of merit is significantly higher at 80 K compared to room temperature. In that case, cryogenic operation will lead significant brightness enhancement under power scaling.
We developed a pulsed Raman laser based on synthetic diamond crystals pumped by a high-brightness Q-switched cryogenic Yb:YAG master oscillator. The maximum output power of 24.5 W was the highest reported for such lasers, with a corresponding slope efficiency of 57%. Beyond the ten-fold improvement of the output performances, the present experiments also provide additional information for the design of high power diamond Raman lasers. We compared the performances of an anti-reflection coated crystal with a Brewster cut sample; the latter had a laser threshold almost 3 times higher due to oblique incidence and shorter optical path length. Raman oscillation was achieved at both room temperature and under cryogenic operation at 77K. The laser threshold was unchanged which indicates that the Raman gain does not vary significantly over that range of temperature. Modeling of these experiments allowed us to confirm the value of Raman gain coefficient of diamond; all sets of data could be fitted with no adjustable parameters, the corresponding Raman gain was 13.5 ± 2.0cm/GW for a pump wavelength of 1030 nm, which agrees with previously published values and the given experimental uncertainty. The present experiment also confirms that the diamond heat-spreader and assembly techniques lead to efficient heat transfer and cooling of the single crystal diamond, and that further power scaling can be achieved.
This material is based upon work supported by the Defense Advanced Research Projects Agency under SPAWAR/SYSCEN Pacific Contract No. N66001-09-C-2079. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of DARPA or SPAWAR/SYSCEN Pacific. This material has been approved for public release with unlimited distribution under case reference 16328.
References and links
1. J. Reintjes, R. H. Lehmberg, R. S. F. Chang, M. T. Duignan, and G. Calame, “Beam cleanup with stimulated Raman scattering in the intensity-averaging regime,” J. Opt. Soc. Am. B 3(10), 1408–1427 (1986). [CrossRef]
2. G. Eckhardt, D. P. Bortfeld, and M. Geller, “Stimulated emission of Stokes and anti-Stokes Raman lines from diamond, calcite, and a-sulfur single crystals,” Appl. Phys. Lett. 3(8), 137–138 (1963). [CrossRef]
3. J. R. Olson, R. Pohl, J. Vandersande, A. Zoltan, T. R. Anthony, and W. F. Banholzer, “Thermal conductivity of diamond between 170 and 1200K and the isotope effect,” Phys. Rev. B 47(22), 14850–14856 (1993). [CrossRef]
4. T. Ruf, M. Cardona, C. S. J. Pickles, and R. Sussmann, “Temperature dependence of the refractive index of diamond up to 925K,” Phys. Rev. B 62(24), 16578–16581 (2000). [CrossRef]
5. I. Friel, S. L. Clewes, H. K. Dhillon, N. Perkins, D. J. Twitchen, and G. A. Scarsbrook, “Control of surface and bulk crystalline quality in single crystal diamond grown by chemical vapour deposition,” Diamond Related Materials 18(5-8), 808–815 (2009). [CrossRef]
6. G. Turri, Y. Chen, M. Bass, D. Orchard, J. E. Butler, S. Magana, T. Feygelson, D. Thiel, K. Fourspring, R. V. Dewees, J. M. Bennett, J. Pentony, S. Hawkins, M. Baronowski, A. Guenthner, M. D. Seltzer, D. C. Harris, and C. M. Stickley, “Optical absorption, depolarization, and scatter of epitaxial single-crystal chemical-vapor-deposited diamond at 1.064μm,” Opt. Eng. 46(6), 064002 (2007). [CrossRef]
7. R. P. Mildren, J. E. Butler, and J. R. Rabeau, “CVD-diamond external cavity Raman laser at 573 nm,” Opt. Express 16(23), 18950–18955 (2008). [CrossRef]
10. W. Lubeigt, G. M. Bonner, J. E. Hastie, M. D. Dawson, D. Burns, and A. J. Kemp, “An intra-cavity Raman laser using synthetic single-crystal diamond,” Opt. Express 18(16), 16765–16770 (2010). [CrossRef] [PubMed]
13. A. A. Kaminskii, R. J. Hemley, J. Lai, C. S. Yan, H. K. Mao, V. G. Ralchenko, H. J. Eichler, and H. Rhee, “High-order stimulated Raman scattering in CVD single crystal diamond,” Laser Phys. Lett. 4(5), 350–353 (2007). [CrossRef]
14. T. T. Basiev, A. A. Sobol, P. G. Zverev, V. V. Osiko, and R. C. Powell, “Comparative spontaneous Raman spectroscopy of crystals for Raman lasers,” Appl. Opt. 38(3), 594–598 (1999). [CrossRef]
15. D. J. Ripin, J. R. Ochoa, R. L. Aggarwal, and T. Y. Fan, “300-W cryogenically cooled Yb:YAG laser,” IEEE J. Quantum Electron. 41(10), 1274–1277 (2005). [CrossRef]
16. T. Y. Fan, D. J. Ripin, R. L. Aggarwal, J. R. Ochoa, B. Chann, M. Tilleman, and J. Spitzberg, “Cryogenic Yb3+-doped solid-state lasers,” IEEE J. Sel. Top. Quantum Electron. 13(3), 448–459 (2007). [CrossRef]
17. J. K. Brasseur, A. K. Abeeluck, A. R. Awtry, L. S. Meng, K. E. Shortoff, N. J. Miller, R. K. Hampton, M. H. Cuchiara, and D. K. Neumann, “2.3-kW continuous operation cryogenic Yb:YAG laser,” Proc. SPIE 6952, 69520L, 69520L-8 (2008). [CrossRef]
19. V. A. Lisinetskii, T. Riesbeck, H. Rhee, H. J. Eichler, and V. A. Orlovich, “High average power generation in barium nitrate Raman laser,” Appl. Phys. B 99(1-2), 127–134 (2010). [CrossRef]
20. J. G. Manni, J. D. Hybl, D. Rand, D. J. Ripin, J. R. Ochoa, and T. Y. Fan, “100-W Q-switched cryogenically cooled Yb:YAG laser,” IEEE J. Quantum Electron. 46(1), 95–98 (2010). [CrossRef]
21. J. Mossbrucker and T. Grotjohn, “Determination of local crystal orientation of diamond using polarized Raman spectra,” Diamond Related Materials 5(11), 1333–1343 (1996). [CrossRef]
22. R. Radebaugh, “Thermal conductance of indium solder joints at low temperature,” Rev. Sci. Instrum. 48(1), 93–94 (1977). [CrossRef]
23. J. C. Diels, and W. Rudolph, Ultrashort Laser Pulse Phenomena (Elsevier, 2006, 2nd edition), pp 328–332.
24. J. K. Brasseur, P. A. Roos, K. S. Repasky, and J. L. Carlsten, “Characterization of a continuous-wave Raman laser in H2,” J. Opt. Soc. Am. B 16(8), 1305–1312 (1999). [CrossRef]
25. G. Boyd, W. Johnston, and I. Kaminow, “Optimization of the stimulated Raman scattering threshold,” IEEE J. Quantum Electron. 5(4), 203–206 (1969). [CrossRef]
26. N. Bloembergen, Nonlinear Optics (Benjamin, 1965) Chap. 5.
27. A. A. Kaminskii, V. G. Ralchenko, and V. I. Konov, “Observation of stimulated Raman scattering in CVD-diamond,” JETP Lett. 80(4), 267–270 (2004). [CrossRef]
28. S. A. Solin and A. K. Ramdas, “Raman spectrum of diamond,” Phys. Rev. B 1(4), 1687–1698 (1970). [CrossRef]
29. The CVD diamond booklet, available at http://www.diamond-materials.com/download
30. W. Koechner, Solid-State Laser Engineering (Springer, 1999) Chap. 7.