Average power and brightness scaling of diamond Raman lasers

Most important research results that explain why the work was done, what was accomplished, and how it pushed scientific frontiers or advanced the field. Diamond holds substantial promise as a high Raman gain laser material with outstanding power handling capability, yet despite this the highest reported output power from a diamond Raman laser prior to this project was approximately 1 W. This report describes investigations into order‐of‐magnitude power scaling of diamond Raman lasers. The investigations focus on 1064 nm beam conversion of 50 W lasers in the “external cavity” Raman cavity configuration in both pulsed and continuous wave modes of operation. For pulsed operation, output powers up to 16 W are demonstrated with 40‐50% conversion efficiency from a compact acousto‐optically Q‐switched neodymium pump laser. The output power is similar to the highest achieved by other groups (in diamond and other Raman materials) but with much greater efficiency and using a much simpler overall system. More than 13 W is also demonstrated at the second‐Stokes wavelength at 1.49 m in the so‐called eye‐safe region, a result which compares well in terms of efficiency as well as power with alternative leading high pulse rate eye‐safe laser technologies. The model‐backed experiments suggest thermal effects are negligible at the current power levels, and much higher output powers are likely when using high power pumps. For continuous wave operation, output powers up to 16 W are demonstrated with conversion efficiency up to 40%. Efficiencies more closely approaching the quantum limit (86%) may be enabled by future improvements in diamond quality. As far as we are aware these are the first demonstrations of CW Raman conversion in a discrete all‐solid‐state system at multi‐watt powers. We show that the concept is applicable for output powers up to several hundred watts and holds promise as a widely applicable technology for high power CW conversion. We also report findings of a UV etching mechanism for diamond surfaces. This multi‐photon effect will need to be considered when using short wavelength pumps but may also provide a novel direct‐write method for creating high resolution structures such as laser waveguides. Introduction: Summary of specific aims of the research and describe the importance and ultimate goal of the work. Although diamond is well known as an excellent Raman material, the emergence of diamond Raman lasers as practical devices has emerged only recently with the production of low absorption Type IIa diamond using the growth method of chemical vapor deposition. Prior to the start of this project in 2009, my team had first demonstrated synthetic diamond Raman lasers and reported conversion efficiencies higher than that observed in any other crystalline material. This initial work was performed at average output powers of approximately 1.2‐Watts. This project aims to leverage diamond’s extraordinary thermal properties to demonstrate Raman laser devices at much higher average powers. The specific aims are to: • Assess in detail of the potential for diamond Raman lasers to provide wavelength and brightness conversion of mature one micron laser technologies. • Establish design rules for scaling towards kilowatt power levels. These aims were addressed by investigating pulsed and continuous wave diamond Raman laser systems using Nd laser pumps at powers up to 50 W. Power scaling and limiting factors were investigated by analyzing system performance and accompanying model simulations. This work is important in that it explores an active laser material with the highest, by a large margin, known room temperature thermal conductivity as well as a host of other thermal advantages. The research is therefore of potentially fundamental importance for developing compact high‐power wavelength and brightness converters that minimize deleterious thermal effects on important performance metrics such as efficiency and brightness. The resulting devices may provide major advantages in wavelength‐choice and/or brightness over alternative technologies such as optical parametric oscillators and Raman shifters based on non‐diamond materials, and comprise a suitable platform for Raman beam combination with high average output power. Experiment: Description of the experiment(s)/theory and equipment or analyses. The experiments can be conveniently divided into two streams of investigation: Pulsed and continuous wave (CW) diamond Raman lasers (DRLs). In both cases, the external cavity Raman laser architecture, in which the Raman laser is pumped by a beam generated separately, was the favored approach as it is most applicable to standard pulsed pump laser such as commercially available Nd or Yb laser systems. The initial proposal had planned to investigate the more restrictive intracavity architecture as a contingency in the case that the external cavity conversion was unsuccessful; however, the latter approach proved to work well and this contingency was not needed. Along with these main streams, work was carried out to characterize the quality and measure pertinent optical properties of the as‐supplied diamond material. Modeling of the Raman lasers was also performed to underpin the experiments and provide a tool for predicting performance when increasing output power beyond the scope of this preliminary study. This parallel work is summarized in the Results and Discussion. 1) Power scaling of pulsed DRLs The basic optical arrangement for the pulsed external cavity Raman lasers is shown in the figure below. The pump laser was based on a commercial side‐diode‐pumped neodymium vanadate gain modules (Northrop Grumman RBA20‐1C2‐FR1‐1013) and Q‐switched using an acousto‐optic modulator to generate pulses approximately 20 ns in duration at 30‐40 kHz repetition rate. The average output power was up to 50 W. The maximum peak power was approximately 50 kW. The intracavity architecture takes advantage of the resonantly enhanced fundamental field to lower the Raman threshold and to enable low threshold CW/quasi‐CW operation. The collimated pump beam was focused into the diamond crystal so that the confocal parameter of the pump field closely matched that of the diamond length. The diamond was up to 9.5‐mm long, and made of low‐nitrogen content, ultra‐low birefringence, Type IIa single crystal diamond grown by chemical vapor deposition and supplied by Element 6 (UK). Anti‐reflection coated and Brewster cut crystals were investigated. The output power, conversion efficiency and beam properties of the DRL were investigated as functions of the input beam properties and Raman cavity mirrors. The experiments have been detailed in Refs [4,15,17] and the results summarized in [11,12,19]. 2) Power scaling of CWDRLs The CW experiments involve the same basic approach but with careful attention paid to cavity losses and to the waist sizes of the pump and Stokes beam in the diamond crystal. Much smaller spot‐sizes in the Raman crystal to are used to enable efficient conversion at the typically much lower peak powers of CW operation. The Raman resonator is nearly concentric to provide a small resonated Stokes mode in the Raman crystal. The resonator output coupling is typically less than 1% and the diamond crystals are selected for low absorption loss and low birefringence. The experiments have been detailed in [1,7,9,10] and summarized in [3,6,11, 12,19]. Results and Discussion: Experimental and/or theoretical research advances or findings and their significance to the field and what work may be performed in the future as a follow on project. 1) Pulsed We find that the efficiencies routinely seen at the 1 W output power level are sustainable at powers up to at least 15 W of Raman output power. The system investigated in most detail consisted of a 9.5‐mm‐long anti‐reflection coated diamond slab placed in 2.5‐cm long external‐cavity Raman resonator and pumped by a 35 W Q‐switched Nd:YVO4 laser generating 22 ns pulses at a 36.5 kHz pulse repetition frequency (preliminary aspects of this work have been reported in [4] attached). The Raman laser mirror set enabled combined first (1240 nm) and second Stokes (1485 nm) output with powers up to 14.5 W at a conversion efficiency of 48.5% (the incident power was 30 W). At maximum power, 13 W was generated at 1485 nm in the eye‐safe spectral region. The total output power is comparable to recent reports in barium nitrate (17 W, Chulkov et al 2012) and diamond (combined two‐beam output of 24.5 W, Feve et al 2011), but with much higher efficiency and using a much simpler overall system. The quantum conversion efficiency is 2.1 times the 17 W barium nitrate and 4.5 times the 24.5 W diamond lasers. The higher efficiency, and the use of a simple and compact high power pump laser compared to these previous studies, is of fundamental importance for applications. Indeed the large fraction of output at 1485 nm, along with the high pulse repetition rate (36.5 kHz) is expected to be of interest for applications of current importance such as rapid scanning remote detection and environmental sensing. The eye‐safe power (13 W) is similar to the highest power achieved using high pulse rate the optical parametric oscillators (13.6 W, Dong et al 2009), and in‐band pumped Er:YAG lasers (9.5 W, Setzler et al 2005) which comprise the most prominent 1.5 m eye‐safe laser technologies (see Table below). No optical damage or deleterious thermal effects were observed. The power characteristics agree well with simple rate equation modeling without the need for including thermal effects. We concluded that further increases in power and efficiency are likely to be achieved without the use of thermal lens or birefringence compensation. By using an upgraded 50 W pump laser we have recently observed output powers at 16.1 W at a single Stokes order at 40% conversion efficiency. Our future pulsed investigations will be aimed towards increasing first Stokes and eye‐safe laser output power by optimizing the output coupling and by using higher power pumps. Based on literature values for key diamond’s thermo‐optical properties, which have been summarized by the Principal Investigator in [18], the thermal limiting factors of thermal lensing, thermally‐induced stress birefringence and stress fracture, will not come into play until the kilowatt level. Optical damage to the diamond facets, which was found to limit power in the work of Feve et al, is an important consideration when power scaling. We observed damage free operation routinely for incident Q‐switched pulse intensities of approximately 300 MW.cm and similar circulating intracavity first Stokes intensity. For a second Stokes laser, damage was observed to the anti‐reflection coatings when using an output coupler of relatively high reflectivity (R = 60% at the second Stokes) and with estimated cumulative (pump plus Stokes) intensities of approximately 1 GW.cm. A further important aspect of this work concerns brightness conversion. We have studied beam quality and efficiency for strongly aberrated pumps, the details of which we will disclose in the near future. Major conclusions: 1) Pulsed DRLs of maximum power up to 16.1 W have been demonstrated, a 13‐times increase over the highest power reported at project start. 2) Conversion efficiencies 40‐50% were obtained at this power level (cf., the record for a crystal Raman laser is 64%). 3) Second‐Stokes conversion of 1.06 m pumps to 1.5 m was demonstrated. The system approach is a promising competing technology for high power (multi‐kilohertz) eye‐safe lasers. 4) Lack of optical damage and deleterious thermal effects portends good prospects for achieving much higher output by using higher power pumps. Thermal effects in the diamond are expected to become significant at kilowatt power levels. In the work of Chulkov et al, a 60 W Q‐switched‐burst flashlamp‐pumped system was used as the pump source and substantial efforts were put into compensation of the induced Raman lens. For Feve et al, the pump source comprised of a cryogenic Yb:YAG laser of average power more than 180 W. 2) Continuous wave CW conversion has been demonstrated at output powers up to 16 W. An anti‐reflection coated slab (9‐mm long) was placed in a confocal high‐finesse resonator to create an intense Stokes waist within the crystal. For an output coupling of approximately 0.4%, the laser threshold was 10 W of incident power and a maximum output power of 16 W was obtained for 40 W of input power (40% conversion efficiency). This result represents the first external cavity crystal Raman laser pumped using a CW Nd laser at 1064 nm. The output power is more than three times the highest for a CW crystalline Raman laser, that being for a 5.1 W intracavity diamond Raman laser at 1240 nm (Savitski et al 2012). A detailed description and characterization of a 10 W device has been reported in [1]. The output power increased linearly with input power with slope efficiency 50% as shown inset. As in the pulsed converter, no thermal effects were observed suggesting good prospects for maintaining the slope efficiency with more powerful pumps. This is consistent with theoretical predictions. Clean‐up of the pump beam was observed for experiments involving a pump with beam quality factor M=1.7. The external cavity approach has important benefits for operation at elevated powers. Placement of the Raman crystal at the midpoint of a concentric resonator mitigates the effects of any thermally induced refractive index gradients. Furthermore, it is more straightforward to decouple the thermal lens effects in the pump laser from the Raman laser compared to intracavity designs. The slope and overall conversion efficiency (50% and 32% respectively) are lower than we typically observed for pulsed systems (65‐85% and 40‐50%). The configuration was convenient for investigating system losses since the experiment allowed good access to the unconverted pump laser beam. From the pump depletion, representing the difference between pump and unconverted pump upon alignment of the Raman laser cavity, we deduced the total power coupled into the Stokes and phonon fields. The results are summarized in the inset figure above for operation at maximum power. It was concluded that significant power (approximately 8 W) is lost in the bulk diamond due to absorption and/or scatter. Thus in contrast to the pulsed diamond Raman lasers described above, improved crystal quality is expected to markedly increase the conversion efficiency and output powers. Using the known output coupler transmission (0.4±0.1%) and Stokes power circulating in the Raman cavity, the combined absorption and scatter coefficient of the diamond was deduced to be approximately 0.17±0.05%/cm at 1240 nm, which is close to the absorption coefficient value (approximately 0.1%/cm at 1064 nm) for the diamond supplier’s low nitrogen (20 ppb) material. Lower loss (i.e., lower nitrogen) diamond is predicted to enable up to 18 W output power at 60% conversion efficiency. Characterization of diamond birefringence in the beam direction is found to vary significantly within and across samples. Selection of low birefringence crystals may be an important factor affecting threshold and efficiency, and we plan to quantitatively investigate this aspect in the near future. Our calculations suggest that the thermal lens strength will be the major consideration over birefringence, end face curvature and stress fracture. On this assumption, we have calculated resonator stability as a function of heat deposited in the crystal. For the above specific arrangement with 11.9 W of power deposited in the Stokes beam waist of 31 m, and we calculate that we are about 3‐4 times below the stability limit. Considering also the freedom to reduce the lens strength by increasing the spot‐size in the diamond, we deduce that output powers of up to several hundred watts are possible without major changes in basic design. In contrast, the power limits for the much thermally less‐amenable molecular ion crystalline Raman lasers is two‐orders of magnitude lower. We believe these advances in CW beam conversion are a significant step towards high power CW lasers that can address wavelength specific applications near 1.2 m and 1.5 m, and at their harmonics by use of intra‐ and/or extra‐cavity nonlinear frequency generation. The latter is an important advantage over fiber Raman laser technologies. The scheme has the practical advantage of building on very mature high power CW Nd and Yb laser technologies, and potentially other sources of high power CW output such as chemical lasers and alkali vapor lasers. The concept is applicable to mature and powerful 532 nm lasers currently available (in which case the threshold will reduce due to a Raman gain coefficient at the shorter wavelength), in order to generate red‐yellow output in the ultraviolet by further nonlinear () frequency generation. The concepts are expected to have commercial value and hence we have taken steps to protect the relevant IP. The only CW pumped external cavity crystal Raman laser reported previously as far as we are aware was for a 6.8 cm‐long barium nitrate crystal and using a gas pump laser in the visible (Ar; 514 nm) (Grabtchikov et al 2005) where the gain is high (approximately 50 cm/GW). For this device the efficiency was low (0.16 W for 5.5 W of pump power), and even at this power level thermal lensing effects in the barium nitrate were noted. Low gain materials having relatively low thermal conductivity (i.e., molecular ion Raman crystals) therefore have the critical problem that the output is restricted to powers not much higher than the laser threshold. Diamond’s combined high gain and thermal conductivity presents an opportunity to investigate CW operation using near‐infrared laser pumps and at much higher output powers. Major conclusions: 1) CW external cavity DRLs of maximum power up to 16 W have been demonstrated at 40% conversion efficiency. 2) Beam clean‐up from a pump of M=1.7 to an M=1.1 output beam has been demonstrated at the 10W level. 3) Crystal quality is critical in this application and large (approximately two‐fold) increases in conversion efficiency are possible if lower loss diamond becomes available. 4) Lack of optical damage and deleterious thermal effects portends good prospects for achieving up to several hundred watts by using higher power pumps without major design changes. 5) The system is amenable for high power visible and ultraviolet generation via intracavity and/or extracavity  nonlinear frequency generation. Other outcomes UV surface etching As an unexpected outcome of our early investigations into laser damage of diamond crystal facets, we observed UV‐induced slow etching of diamond surfaces at a rate that increased with the square of the incident fluence up to the ablation threshold. Ablation threshold is taken here to mean the minimum fluence to produce the irregular shaped pit that is characteristic of conventional diamond ablation in the nanosecond regime. For example, for 266 nm pulses of fluence corresponding to 60% of the ablation threshold, pits 400 nm deep were formed after of 30 s exposure at 7 kHz (2.2×10 pulses; corresponding to an etch rate of 2 pm/pulse). For lower fluences, the time required to achieve the same depth was much longer in inverse proportion to the square of the intensity. In each case, the average number of atoms removed per pulse is less than an atomic layer. X‐ray surface studies have shown that the etched surfaces were oxygen terminated and free of graphite. Moreover, the results to date suggest the etch rate is proportional to the two‐photon absorption rate largely irrespective of the surface facet direction, pulse duration in the range 0.02 to 100 ns, and pulse repetition rate. This work has been reported in detail in [2] and was highlighted in Nature News. Lens strength scales with the inverse square of the beam size for a given power deposited. http://www.nature.com/news/2011/110715/full/news.2011.421.html The details of the UV absorption in the surface layers and subsequent ejection kinetics are not well understood. A separate group had previously seen signs of this effect (Kononenko et al 2007) and reported that the presence of oxygen is required for etching. An improved understanding of the mechanism is needed to determine how the etching can be prevented or exploited. It is not yet known whether 3‐photon or higher‐order multiphoton effects lead to etching and at what maximum wavelength the effect becomes negligible. The etching, which can be performed in air that and is well suited for rapid prototyping of direct‐write structures, is also of interest as a novel method for machining of smooth high resolution surface structures in diamond such as laser waveguides. The inset shows, for example, a profile of our University logo of dimensions approximately 100 m by 100 m etched into the {100} face of a single crystal diamond. The square‐dependence on the intensity is reproduced spatially, so that features can be written smaller than the diffraction‐limited spot‐size of the writing beam. We aim to investigate the method further to determine whether the etching is able to solve current challenges in the production of long low‐loss waveguides. Overall assessment Diamond Raman laser technology is highly promising for high power beam conversion and further advances in performance are expected as our research efforts continue. Increases in power and wavelength range are expected in the short term with the currently available material. Parallel advances by crystal growers will also fuel progress, particularly for CW converters in which losses are relatively much more important. With major breakthroughs in the growth of quality large single crystals occurring only in the last decade, it can still be considered early days for synthesis. Reduced absorption and scatter, larger sizes and lower production costs, are currently some directions of development that will assist Raman laser development. Reproducibility and lower cost are also directions that will benefit research and crucial for future transfer of the technology into the commercial arena. Currently CVD growth is perceived to be the most promising method due to its unrivalled level of control on absorbers and residual stress‐birefringence. Studies on other sources of material are few as far as we are aware, however, there may be scope for other types of material (e.g., HPHT or polycrystalline material) at least in some configurations where birefringence or scatter loss are not so critical. Longer crystals than the up‐to‐1 cm‐lengths currently investigated may also provide advantages for increasing performance range. Doping of diamond with laser ions represents a highly attractive avenue of development following the success of Nd doped Raman crystals in simplifying intracavity Raman laser design (“self‐Raman lasers”). However, the present challenges involved in incorporating a sufficient density of lasant ions into the closely packed diamond lattice are large and a major advance is required before this possibility can be exploited. Diamond waveguides are also an area of current development by us and others that will assist device design via the well‐known benefits afforded by beam confinement for enhancing nonlinearity and mitigating thermal effects. Such developments in diamond material engineering create exciting prospects for the field of diamond lasers. Future development of diamond lasers will be driven by the ever increasing demand for laser devices of enhanced performance range. Our future work is aimed at demonstrating a class of devices for wavelength conversion at the power level usually the domain of fiber or disk lasers and offering a greater range of wavelengths and intrinsically narrower output bandwidths compared to fibers. There are potentially a large range of applications demanding wavelength conversion that may benefit from the high power handling capability offered by diamond. List of Publications and Significant Collaborations that resulted from the AOARD supported project: a) papers published in peer-reviewed journals, [1] Kitzler, O., McKay, A. & Mildren, R.P., 2012. “Continuous‐wave wavelength conversion for high‐power 50 m applications using an external cavity diamond Raman laser,” Optics Letters, 37(14), pp.2790–2. [2] Mildren, R.P., Downes, J E, Brown, J D, Johnston, B F, Granados, E, Spence, D J, Lehmann, A Weston, L, Bramble, A., 2011. “Characteristics of 2‐photon ultraviolet laser etching of diamond,” Optical Materials Express, 1(4), pp.576–585. b) papers published in peer-reviewed conference proceedings, [3] Mildren, R.P., 2012. “Recent progress in diamond Raman lasers,” Mater. Res. Soc. Symp. Proc., 1395, pp.1–12. *Invited Plenary Session* c) papers published in non-peer-reviewed journals and conference proceedings, [4] A. McKay, O. Kitzler, H. Liu, D. Fell, and R.P. Mildren."High average power (11 W) eye‐safe diamond Raman laser."Photonics Asia, Conference on High Power Lasers and Applications 8551 (International Society for Optics and Photonics) pp. 85510U‐85510U, (2012). d) conference presentations without papers, [5] O. Kitzler, A. Sabella, A. McKay and R.P. Mildren, “Characterization of optical quality single crystal diamond for Raman laser applications,” XX INTERNATIONAL MATERIALS RESEARCH CONGRESS, 14 ‐ 19 August, Cancún, Mexico, Paper S17‐P05 (2011) Poster Paper [6] R.P. Mildren, “Performance extension of Raman lasers using synthetic diamond,” CLEO Pacrim (Sydney) Aug 28‐Sep1 Paper 3230‐CT‐1 (2011) *Invited Paper* [7] O. Kitzler, A. McKay and R.P. Mildren, “CW diamond laser architecture for high power Raman beam conversion,” IQEC/CLEO Pacific Rim (Sydney) Aug 28‐Sep1 (2011) *Postdeadline paper* [8] O. Kitzler, A. Sabella, B.F. Johnston, A. McKay and R.P. Mildren, “Design and characterization of optical quality synthetic diamond for Raman laser applications,” IQEC/CLEO Pacific Rim (Sydney) Aug 28‐Sep1 Paper 4700‐PO‐20 (2011) [9] O. Kitzler, A. McKay and R.P. Mildren, “High average power diamond Raman beam conversion,” 2012 MMI‐Harvard Diamond Photonics Symposium, (Melb, 17‐20 Jan), P8, (2012). [10] O. Kitzler, A. McKay and R.P. Mildren, “High power CW diamond Raman laser: Analysis of efficiency and parasitic loss,” Conference on Lasers and Electro‐Optics (6 ‐ 11 May 2012 : San Jose, CA) paper CTh1B.7 (2012). [11] R.P. Mildren, O. Kitzler, A. M. McKay and H. Liu, “High‐power Beam Conversion Using Synthetic‐diamond Raman Lasers” High Power Laser and Applications (HPLSA), Istanbul, Paper S11‐8 (2012) [12] R.P. Mildren, O. Kitzler, A. McKay, H. Liu, “Is diamond a good material for wavelength conversion at high power?,” Photonics Asia, High‐Power Lasers and Applications VI, Conference 8551, 5 ‐ 7 November, Paper 8551‐20, (2012) [13] R.P. Mildren, A. Lehmann, C. Baldwin and J.E. Downes, “Polarization Dependent Nanostructuring of Diamond Surfaces by Two‐Photon Ultraviolet Etching,” Australian Institute of Physics Congress, UNSW Dec 9‐13 (2012) [14] A. Sabella, J.A. Piper and R.P. Mildren “Impact of pump polarisation and linewidth on the Raman gain coefficient of diamond”, Australian Institute of Physics Congress, UNSW Dec 9‐13 (2012) [15] A. McKay, O. Kitzler, H. Liu, D. Fell and R.P. Mildren, “Efficient High‐Power Pulse Diamond Raman Laser Operating in the Eye‐Safe Spectral Region” Australian Institute of Physics Congress, UNSW Dec 9‐13 (2012) [16] O. Kitzler, A. McKay and R.P Mildren, “Continuous Wave, 10 W External Cavity Raman laser: Experiment and Modeling,” Australian Institute of Physics Congress, UNSW Dec 9‐13 (2012) e) manuscripts submitted but not yet published, and [17] A. McKay, H. Liu, O. Kitzler, R.P. Mildren, “Efficient 14.5 W diamond Raman laser at high pulse repetition rate with first (1240 nm) and second (1485 nm)Stokes output,” Laser Physics Letters, submitted 28 Dec 2012, (2013) [18] R.P. Mildren, “Intrinsic optical properties of diamond,” in Optical Engineering of Diamond (Wiley) pp1‐ 34, (2012) (Expected release date March 2013) [19] R.P. Mildren, A. Sabella, O. Kitzler, D.J. Spence and A. McKay, “Diamond Raman laser design, performance and prospects,” in Optical Engineering of Diamond (Wiley) pp239‐276,(2012) (Expected release date March 2013) f) provide a list any interactions with industry or with Air Force Research Laboratory scientists or significant collaborations that resulted from this work. Some of the outcomes of this project have led to a successful 3‐year project in ultrafast Raman lasers, jointly funded by industry partner M‐Squared Lasers Ltd (UK) and the Australian Research Council). The scope of this project is heavily concentrated on using diamond for wavelength conversion of laser pumps of pulse duration less than 30 ps at multi‐watt powers. The innovations related to high power CW conversion in an external diamond resonator have also stimulated a number of interactions with companies. Once contract agreements are in place, there is good potential for at least two of these to form formal collaborations. References Chulkov et al., 2012. Thermal aberrations and high power frequency conversion in a barium nitrate Raman laser. Applied Physics B, 106(4), pp.867–875. Chen X.H. et al., 2011. Highly efficient double‐ended diffusion‐bonded Nd:YVO4 1525‐nm eye‐safe Raman laser under direct 880‐nm pumping. Applied Physics B, 106(3), pp.653–656. Dong X.‐L. et al., 2009. High‐power 1.5 and 3.4 μm intracavity KTA OPO driven by a diode‐pumped Q‐switched Nd:YAG laser. Optics Communications, 282, pp1668–1670. Feve J.‐P. M. et al., 2011. High average power diamond Raman laser. Optics Express 19(2), pp913–922. Setzler D. et al., 2005. Resonantly pumped eyesafe erbium lasers. IEEE Journal of Selected Topics in Quantum Electronics. 11(3), 645–657.