Efficient high-power operation of Cr:LiSAF, Cr:LiSGAF, and Cr:LiCAF lasers pumped by broad-area laser diodes is demonstrated. A maximum slope efficiency of 51% and output power of 0.55 W was reached at 1.2 W of absorbed pump power, which is the highest output power to date with broad-area laser diode pumping. With the laser design used the onset of thermal quenching in Cr:LiSAF due to high temperatures was pushed to higher pump powers and good mode matching was achieved.
© 2003 Optical Society of America
Diode-pumped Colquiriite lasers are attractive sources of broadband tunable radiation in the near infrared. Most attention has been given to these crystals in order to produce femtosecond laser pulses [1–3], or to amplify laser pulses in regenerative amplifiers [4–6]. To date the Colquriites exhibit the broadest amplification bandwidth of diode-pumped laser crystals ranging from 110 nm to 220 nm in the near infrared. Additionally, the broad absorption band of 100 nm centered around 650 nm with two maxima at 640 nm and 670 nm, respectively, makes temperature tuning of the pumping laser diodes completely unnecessary in comparison to, e.g., Nd:YAG. The long upper state lifetime of 67 µs for Cr:LiSAF and up to 170 µs for Cr:LiCAF  make these crystals prominent gain media for cw-pumped regenerative amplifiers. However, limitations in scaling the output power arise mainly due to thermal quenching of the upper state lifetime due to excessive heat inside the pumped region, which drastically reduces the available gain . Our report demonstrates how these limitations can be pushed to higher pump power levels. This was done in a cw-configuration to avoid the additional complexity of a regenerative amplifier. Results from the new pumping arrangement apply as well to pulse energy scaling in high-repetitive short-pulse regenerative amplifiers.
The upper state lifetimes of the Colquiriites are strongly temperature dependent. For Cr:LiSAF and Cr:LiSGAF, the lifetime is reduced to half of its room temperature value at a temperature of 69°C and 88°C, respectively. In contrast, the reduction of the upper state lifetime occurs at a much higher temperature of 255°C for Cr:LiCAF [7,9]. With diode pumping, the non-diffraction limited pump radiation typically has to be tightly focused using a short absorption length to achieve a good matching of the pump to the cavity mode. This results in a high absorbed power per unit area and leads to excessive local heating of the laser crystal. As a consequence, the output power drastically decreases in Cr:LiSAF and Cr:LiSGAF. In order to reduce the absorbed power per unit area, we used crystals with a lowered Cr3+-doping concentration to distribute the pump power over a larger absorption length. Furthermore, we used an elliptical pump beam matched to an elliptical cavity mode, a concept first proposed by Krausz et al. . Using an elliptically pumped volume, the generated heat was efficiently removed in a fashion of a one-dimensional heat flow using a thin slab crystal and cylindrical cavity optics, as demonstrated by Kopf et al.  when pumping a Cr:LiSAF laser with a laser diode bar. In contrast to the scheme applied by the above mentioned authors, we used the intrinsic ellipticity of the cavity mode due to a single Brewster-surface inside the resonator without cylindrical optics. The cavity mode was well matched to the output of 100 µm broad-area laser diodes. Additionally, costly cylindrical intracavity optics with typically less optical quality compared to their spherical counterparts were avoided with this design.
The setup of the laser is shown in Fig. 1. It consisted of a flat–Brewster-cut crystal (height × width × length: 1 mm × 4 mm × 5 mm), a spherical focussing mirror of high reflectivity (R > 99.97%; ROC=-200 mm) and a flat end mirror or output coupler, respectively. The flat side of the crystal was coated highly reflective for the laser wavelength and highly transmissive for the pump wavelength. Two laser diodes (100 µm × 1 µm emitter, fast axis collimation, 670 nm, 500 mW each) were polarization-coupled and focused into the crystal from each side, allowing to apply in total up to 2 W of pump power at 670 nm. The crystal was wrapped in Indium foil, mounted in a copper heat sink and cooled by a Peltier element.
2.2 Thermal quenching
In a first experiment, the laser performance of two Cr:LiSAF crystals of 1.5 at% and 3 at% Cr3+-doping concentration were compared in order to demonstrate the benefit of the lower doping level as Cr:LiSAF is the most sensitive crystal of the three investigated laser media exhibiting the lowest critical temperature for thermal quenching. For each measurement we used the optimum output coupling of 2.2% as shown later.
In this experiment the crystal was pumped with two laser diodes from the flat side only at the smaller focus of the pump where thermal effects therefore appear first. This inhibited masking thermal effects by the pump power from the Brewster-cut side. Both polarization-coupled diodes were turned up simultaneously for equal pump beam diameters. Clearly, a distinct drop in output power at the emission wavelength around 848 nm was noticeable for the 3 at% doped crystal at 600 mW of absorbed pump power and 14 °C heat sink temperature, as depicted in Fig. 2. This drop was observed at a higher pump power level of 650 mW when cooling the heat sink down to 10°C. By using the elliptic pump beam shape in our setup we already achieved an improvement of more than 100% compared to a circular pump spot as used by Balembois et al. . Further improvement was achieved with the lower doping concentration of 1.5 at% as we observed no degradation up to the maximum pump power of 650 mW.
The good mode matching was underlined by the fact that the laser output power using the 1.5 at% doped crystal at a given amount of absorbed pump power was the same in the case of single side pumping as well as pumping with all four diodes, since the effective emitting area of the slow axis of the laser diodes increased with increasing output power. Thus, the effective emitting area at a given amount of power was larger for pumping with two diodes compared to pumping with four diodes. Finally it should be mentioned, that the larger amount of absorbed pump power at 3 at% doping was due to the increased absorption compared to 1.5 at% doping.
The roll over in output power is a good indication for a reasonable maximum pump power limit as seen in Fig. 2. We used this result to estimate a further power scalability with an indirect approach. The crystal temperature was increased by heating the crystal via the Peltier element of the heat sink to determine the roll over of the characteristic curve at elevated temperatures. The crystal had to be heated to 45°C until at maximum pump power the onset of a roll over was noticed. Assuming that the roll over occurred latest at a temperature inside the pumped region near the temperature of thermal quenching of 69°C, the temperature difference of the hottest region inside the crystal to the heat sink temperature was at maximum 24°C at 650 mW of absorbed pump power from one side. From this, a temperature rise of 0.037°C/mW was calculated. Thus, at a heat sink temperature of 16°C, up to 1.4 W of absorbed pump power with higher brightness may be applied. Respectively, it should be possible to pump Cr:LiSAF from both sides with up to 2.8 W. Extrapolation for Cr:LiCAF by taking into account the higher quenching temperature of 255°C a total pump power of up to 13 W from both sides should be possible.
2.3 Optimum output coupling and slope efficiencies
To find the optimum degree of output coupling the maximum output power at full pump power was measured for a series of different output couplers. A plot of the normalized output power versus output coupling is given in Fig. 3.
As can be seen, for Cr:LiSAF and Cr:LiSGAF the optimum output coupling was about 2.2%. Due to the lower spontaneous emission cross-section of Cr:LiCAF  and thus, lower small signal gain at the same pumping level, the optimum coupling was at 1.25%. Also, Cr:LiCAF showed a faster decrease in output power for over-coupling in comparison to Cr:LiSAF and Cr:LiSGAF attributed to the lower gain of Cr:LiCAF.
After the optimum output coupling for each laser material was determined, the laser performance of Cr:LiCAF and Cr:LiSGAF was investigated in comparison to Cr:LiSAF. The doping level of Cr:LiSGAF and Cr:LiCAF was chosen as nominally 2.0 at% and 3.5 at%, such that the effective absorption coefficient was nearly the same (±5%) compared to the 1.5 at% doped Cr:LiSAF crystal used. This resulted in a waste heat generation per unit area due to the absorbed pump power of about an equal amount for all crystals investigated. The output power versus absorbed pump power at optimum output coupling is given in Fig. 4.
All three crystals showed about the same performance. The highest output power of 550 mW at 1.2 W of absorbed pump power and a slope efficiency of 51% was reached with Cr:LiSGAF. The output wavelengths were centered at 840 nm and the threshold pump power was 120 mW. Almost the same output power of 530 mW was reached with Cr:LiSAF at the same amount of pump power and with a slope efficiency of 51%. The threshold pump power was slightly higher with 150 mW and the output wavelength was located at 848 nm. Due to the coating of the Cr:LiCAF crystal, only a maximum of 1.1 W of absorbed pump power was available with this crystal. Because of the coating of the mirrors, the emission wavelength was located at 800 nm rather than at the gain maximum of 765 nm, which resulted in a reduction of the small signal gain of approximately 10%. The emitted maximum power was 440 mW at a slope efficiency of 47% and a threshold pump power of 135 mW. To the best of our knowledge, all above values of output power and slope efficiency are the highest to date for Colquiriite lasers pumped by broad-area laser diodes. For comparison, in Tab. 1 the highest slope efficiencies for pumping with a diffraction limited beam and broad-area laser diodes of selected publications are given together with values of this work.
The efficiency of Cr:LiSGAF in our experiments was within 1% of the value obtained with diffraction limited pumping . The lower slope efficiency of Cr:LiSAF with diffraction limited pumping might be explained by the low crystal quality of the reported high scatter loss of 1.1% . The measured scattering loss of our crystals was 0.14% for Cr:LiSAF and Cr:LiSGAF, respectively, and 0.28% for Cr:LiCAF.
Comparing the efficiency of reference  and  for Cr:LiCAF in Tab. 1 with our own work strongly suggests a parasitic loss as a function of doping concentration in Cr:LiCAF. For verification of this hypothesis we compared experimental results from various groups concerning the used doping level, pump source and output coupling as given in Tab. 2. The slope efficiency of Cr:LiCAF from this work stated in Tab. 2 was measured with a 4.3% output coupler for better comparability.
A typical slope efficiency of approximately 50% was reached in all of the first three works of Tab. 2, regardless of doping level and degree of output coupling. On the contrary, the highest efficiency of 68.2% was reached only with a very low doping level of 0.3 at% ; a sample of 0.94 at% doping level investigated by the same authors already revealed a reduction in slope efficiency to 52%. This suggests a doping level dependent parasitic loss mechanism.
Since the laser inversion clamps at its threshold value for homogeneously broadened media, the (Auger) upconversion process may be excluded from consideration . A process which affects the slope efficiency and is present in all Colquiriite crystals is excited state absorption (ESA) at the laser wavelength [19,20]. In this process, an electron of an excited ion absorbs a photon and is excited to higher lying levels. Consequently, both the photon and the energy of the excited ion are lost for the lasing process. Concluding from the experimental results stated in the Tab. 2, a doping level dependent ESA may be assumed, which reduces the slope efficiency at higher doping levels in Cr:LiCAF, but further related investigations are necessary.
In conclusion, a new practical pumping geometry and cavity design have been demonstrated, which substantially reduce the laser crystal temperature for longitudinally diode-pumped Cr3+-lasers. Without using cylindrical optics an elliptical cavity mode was formed in a flat-Brewster-cut laser crystal. From our measurements, one can deduce that no limitation by thermal quenching will occur for pump power levels up to 2.8 W for Cr:LiSAF when 100 µm diode emitters are used in a two-sided pumping arrangement.
With an available absorbed pump power of 1.2 W of four broad-area diodes a slope efficiency of 51% and an output power of 0.55 W were generated from Cr:LiSGAF, which is to date the highest output power with broad-area diode pumping.
By extrapolating our measurements, in Cr:LiCAF a roll over in output power due to thermal quenching is not expected until a pump power above 13 W at two-sided pump power delivery is reached.
Despite the present excited state absorption in Cr:LiCAF, this laser material shows a promising potential for power scaling, due to lower thermal quenching, whereas for low power applications up to a few hundred milliwatts of output power Cr:LiSAF or Cr:LiSGAF are superior, due to the three times larger stimulated emission cross-section.
Support of the BMBF is gratefully acknowledged under contract 13N7212.
References and Links
2. I.T. Sorokina, E. Sorokin, E. Wintner, A. Cassanho, H.P. Jenssen, and R. Szipöcs, “Prismless passively mode-locked femtosecond Cr:LiSGaF laser,” Opt. Lett. 21 (15), 1165–1167 (1996). [CrossRef] [PubMed]
3. K.M. Gäbel, P. Rußbüldt, R. Lebert, and A. Valster, “Diode Pumped Cr3+:LiCAF fs-Laser,” Opt. Commun. 157, 327–334 (1998). [CrossRef]
4. R. Mellish, N.P. Barry, S.C.W. Hyde, R. Jones, P.M.W. French, J.R. Taylor, C.J. Ven der Poel, and A. Valster, “Diode-pumped Cr:LiSAF all-solid-state femtosecond oscillator and regenerative amplifier,” Opt. Lett. 20 (22), 2312–2314 (1995). [CrossRef] [PubMed]
5. D. Parsons-Karavassilis, R. Jones, M.J. Cole, P.M.W. French, and J.R. Taylor, “Diode-pumped all-solid-state ultrafast Cr:LiSGAF laser oscillator-amplifier system applied to laser ablation,” Opt. Commun. 175, 389–396 (2000). [CrossRef]
6. A. Isemann, H. Hundertmark, and C. Fallnich, “Diode-pumped Cr:LiCAF fs regenerative amplifier system seeded by an Er-doped mode-locked fiber laser,” Appl. Phys. B 74, 299–306 (2002). [CrossRef]
7. L.K. Smith, S.A. Payne, W.L. Kway, L.L. Chase, and B.H.T. Chai, “Investigation of the laser properties of Cr3+:LiSrGaF6,” IEEE J. Quantum Electron. 28 (11), 2612–2618 (1992). [CrossRef]
8. F. Balembois, F. Druon, F. Falcoz, P. Georges, and A. Brun, “Comparison between Cr:LiSAF and Cr:LiSGAF for cw diode-pumped Q-Switch operation,” OSA TOPS Advanced Solid State Lasers 10, of OSA Proceedings Series (Optical Society of America, D.C., 1997), pp. 170–173.
9. M. Stalder, M. Bass, and B. Chai, “Thermal quenching of fluorescence in chromium-doped fluoride laser crystals,” J. Opt. Soc. Am. B 9 (12), 2271–2273 (1992). [CrossRef]
11. D. Kopf, J. Aus der Au, U. Keller, G.L. Bona, and P. Roentgen, “400-mW continuous-wave diode-pumped Cr:LiSAF laser based on a power-scalable concept,” Opt. Lett. 20 (17), 1782–1784 (1995). [CrossRef] [PubMed]
12. F. Balembois, F. Falcoz, F. Kerboull, F. Druon, P. Georges, and A. Brun, “Theoretical and Experimental Investigation of Small-Signal Gain for a Diode-Pumped Q-Switched Cr:LiSAF Laser,” IEEE J. Quantum Electron. 33 (2), 269–278 (1997). [CrossRef]
13. S. A. Payne, L. Chase, and G. Wilke, “Optical spectroscopy of the new laser materials, LiSrAlF6:Cr3+ and LiCaAlF6:Cr3+,” J. Lumin. 44, 167–176 (1989). [CrossRef]
14. L. K. Smith, S. A. Payne, W. L. Kway, L. L. Chase, and B. H. T. Chai, “Investigation of the laser properties of Cr3+ :LiSrGaF6,” IEEE J. Quantum Electron. 28 (11), 2612–2618 (1992). [CrossRef]
15. S. A. Payne, L. Chase, H. W. Newkirk, L. K. Smith, and W. F. Krupke, LiCaAlF6:Cr3+: A promising new solid-state laser material,” IEEE J. Quantum Electron. 24 (11), 2243–2252 (1988). [CrossRef]
16. R. Scheps, “Cr-doped solid state lasers pumped by visible laser diodes,” Opt. Mat. 1, 1–9 (1992). [CrossRef]
17. F. Loesel, C. Horvath, F. Grasbon, M. Jost, and M. H. Niemz, “Selfstarting femtosecond operation and transient dynamics of a diode-endpumped Cr:LiSGaF laser with a semiconductor saturable absorber mirror,” Appl. Phys. B 65, 783–787 (1997). [CrossRef]
18. I.T. Sorokina, E. Sorokin, E. Wintner, A. Cassanho, H.P. Jenssen, and M.A. Noginov, “Efficient Continuous-Wave TEM00 and Femtosecond Kerr-Lens Mode-Locked Cr:LiSrGaF Laser,” Opt. Lett. 21 (3), 204–206 (1996). [CrossRef] [PubMed]
19. I. Sorokina, E. Sorokin, E. Wintner, A. Cassanho, and H. Jenssen, “In Situ measurement of ESA, upconversion, and thermal quenching in Cr:LiSAF and Cr:LiSGAF lasers,” Advanced Solid State Lasers, OSA Technical Digest (Optical Society of America, Washington, D.C., 1997), pp. 411–414.
20. P. Beaud, M. C. Richardson, Y.-F. Chen, and B. H. T. Chai, “Optical amplification characteristics of Cr:LiSAF and Cr:LiCAF under flashlamp-pumping,” IEEE J. Quantum Electron. 30 (5), 1259–1266 (1994). [CrossRef]