Open Access
Add to BibSonomyAbstract
The laser performance of the monoclinic Ho:RE(WO4)2 (RE = Y, Gd, Lu) crystals is compared under identical experimental conditions. The comparison deals with the laser transition of Ho3+ at ~2.1µm by using two different pump sources, a diode laser operating at 1941 nm and a diode-pumped Tm:KLu(WO4)2 laser operating at 1946 nm. The results show internal slope efficiencies of ~60% and output powers exceeding 400 mW. The laser performance of Ho:KY(WO4)2 and Ho:KLu(WO4)2 is quite similar and superior to Ho:KGd(WO4)2.
©2011 Optical Society of America
1. Introduction
Currently, there is an increasing interest in infrared solid-state lasers based on Ho3+ ions (Ho) operating slightly above 2 μm due to the potential applications in the fields of medicine, remote sensing and as a pump source for OPO´s [1]. Essential advantages of the Ho-doped materials in comparison to their Tm-doped analogues is that such laser crystals offer higher gain cross-section and longer lifetime of the emitting 5I7 → 5I8 transition near 2.1 µm. The main drawback of the Ho ion is that it does not exhibit suitable absorption bands for conventional laser diodes. In order to overcome this drawback many researchers used Tm as sensitizer with diode pumping around 800 nm [2,3]. However, this co-doped system exhibits up-conversion losses which in turn, lead to poor efficiency. More efficient is the alternative pump scheme using direct pumping of the 5I7 emitting level through Tm lasers operating near 1.9 μm, including Tm crystals and Tm fibers [4–6]. More recently, however, efficient diode lasers operating also at around 1.9 µm have demonstrated to be promising for power scaling of Ho–based lasers [7]. The essential advantage of direct excitation of the Ho emitting level (the so called in-band pumping) is the minimum heat generated in the crystals due to the small quantum defect between the pump and laser wavelengths and the high pump efficiency which are both prerequisites for power scaling of the 2.1 μm laser.
The well known laser hosts, the monoclinic KRE(WO4)2 (RE = Y, Gd, Lu) crystals (hereafter KREW) have proved to be very suitable for Yb and Tm ion doping, for efficient laser operation, including high-power laser designs such as slabs and thin disks [8–11]. This family of laser crystals is characterized by large transition cross-sections, strong polarization anisotropy and weak concentration quenching. Apart from Yb and Tm, we recently demonstrated laser generation in Ho:KLuW laser using a diode-pumped Tm:KLuW laser [12]. While the KLuW host turned out to be most suitable for Yb- and Tm-doping, this issue is still open for the case of Ho-doping. In this work, we compare the laser performance of in-band pumped Ho–doped KREW crystals (the three hosts of the same family) under two different pump sources, a diode laser operating at 1941 nm and a diode-pumped Tm:KLuW laser operating at 1946 nm.
2. Experiment
A series of Ho-doped KREW crystals have been grown at a doping concentration of 3 at.% by the Top Seeded Solution Growth Slow Cooling method (TSSG-SC). The concentration in the crystals was measured using the electron probe micro analysis (EPMA) technique. Due to the strong anisotropy of the monoclinic KREW crystals [11], the samples were cut and polished along the principal optical directions, namely Ng, Nm and Np, associated with the three refractive indices ng, nm and np of this biaxial crystal. The samples were plates with faces perpendicular to the principal optical directions. Spectroscopic characterization of Ho:KREW crystals was performed in terms of polarized room and low temperature optical absorption measured using a Cary Varian 500 spectrophotometer. The emission cross-section of the transition near 2.1 µm was computed by means of the reciprocity method [13] from the absorption cross-section spectra.
Concerning the laser experiment, two types of pump sources were used. One is a fiber-coupled GaSb laser diode module (Dilas, Compact 16/400) which had a maximum output power of 16 W and wavelength varying from 1932 to 1941 nm depending on the current level. The spectral bandwidth (FWHM) increased from 4 nm at the laser threshold to 7 nm at maximum current level. The core diameter of the fiber was 400 μm. A special lens assembly attached to the end of the fiber was used to collimate and focus the unpolarized pump beam with a 2:1 imaging ratio. The second pump source is a diode–pumped Tm:KLuW laser described in a previous work [9] delivering a maximum of 4 W centered at 1946 nm. The laser radiation from the Tm:KLuW laser was linearly polarized parallel to its Nm principal optical axis. Both experiments were performed in a hemispherical two mirror linear cavity (see Fig. 1a and 1b). It was formed by a plane pump mirror (M5), antireflection (AR) coated for the pump wavelengths and high-reflection (HR) coated for the laser wavelength (2000-2100 nm). The output coupler (M6) had a radius of curvature RC = 50 mm. Its transmission (Toc) was 1.5%( ± 0.3)%, 3 ( ± 0.5)%, 5( ± 1.0)%, and 20( ± 4.0)% in the 1820 - 2050 nm range. In the case of diode pumping, the resulting pump spot size was 220 µm in diameter and the expected laser spot size was calculated to be 130 µm. In the case of Tm:KLuW laser pumping, the collimated pump beam was passed through an isolator (I) to avoid any feedback from the Ho-laser cavity to the Tm laser. A half-wave plate after the isolator was used to adjust vertical polarization for pumping the Ho-crystals parallel to the E//Nm principal optical axis to take advantage of the somewhat higher absorption cross section compared to E//Np. The resulting pump spot size in this configuration was measured to be 130 µm in diameter and the calculated laser waist was nearly the same as in the diode-pumped cavity. The cavity length was around 49 mm in both cases.
Fig. 1 Laser setups: a) Diode pumped Ho:KREW laser (L1 (imaging lens assembly consisting of a collimator and aspherical lens with f = 6 mm), M5 (plane pump mirror), M6 (Ho–laser output coupler). b) Tm:KLuW pumped Ho:KREW laser (λ/2 - half-wave plate, I – isolator, L1 (AR-coated lens assembly for collimation and focusing of the fiber output with f = 30 mm), L2 (AR-coated collimating lens with f = 50 mm), L3 (AR-coated focusing lens with f = 50 mm), M3, M4 (45° bending mirrors), M1 (plane pump mirror, AR @ 800 nm and HR for laser wavelength), M2 (Tm-laser output coupler with RC = 25 mm and Toc = 3%), M5 (plane pump mirror) and M6 (Ho-laser output coupler).
The dimensions of the AR-coated active elements (Ng-cut) were 3 × 3 × 3 mm3. They were mounted in a Cu-holder maintained at 16°C for cooling, with indium foil for better contact of the top and bottom surfaces. The samples were located as close as possible to the pump mirror and were positioned under normal incidence to the pump beam.
3. Results and discussion
Crystals of several doping concentrations were grown by the TSSG-SC method as mentioned previously. For growing the single crystals, we introduced 200 g of solution mixture inside the furnace with the binary composition of 12 mol% Ho:KREW as solute and 88 mol% K2W2O7 as solvent in a platinum crucible of 50 mm diameter and 50 mm length using K2CO3, WO3, RE2O3 (RE = Y, Gd, Lu), and Ho2O3 with analytical grade of purity. The mixture was homogenized by maintaining the solution at 50 K above the saturation temperature for 5 to 6 hours. The thermal axial gradient of the furnace was ~1.5 K/cm. The seeding process was realized with a b crystallographic oriented seed (e.g. for growing Ho:KYW crystals, a KYW seed was used) located at the centre of the surface of the solution. After accurately determining the saturation temperature by observing the growth and dissolution of the seed in contact with the surface of the solution, the crystals grew by slow cooling at the rate of 0.1 K/h for ~20 K. The grown crystals were slowly removed from the solution and kept above its surface for cooling down to room-temperature at a rate of 25 K/h, sufficiently to avoid thermal shock. The detailed growth methodology is described elsewhere [14]. During the growth process the seed was rotated at 40 rpm for KYW and KLuW and 60 rpm for KGdW, in order to enhance mass and heat transport in the solution and to avoid inclusions. Since only 3 at.% Ho:KLuW showed laser operation in the previous experiment [12], for a comparative study in this work we considered only 3 at.% Ho:KREW crystals. Table 1 gives a summary of the crystal growth details and Table 2 summarizes the dopant content in the crystals as well as the actual stoichiometric chemical formula.
Table 1. Summary of the Ho:KREW crystal growth experiments
Table 2. Summary of the EPMA (Electron Probe Micro analyzer) results for the grown Ho:KREW crystals
From the optical absorption measurements, we analyzed the absorption features of the 5I7 level. Figure 2 (a, b, c) shows the absorption cross-section for the three hosts in the 1800 – 2200 nm range for E//Nm and Np. Note the high degree of anisotropy where Nm absorbs almost two times more than Np and the broadband spectral features. The maximum of absorption is centered near 1960 nm with a Full-Width at Half Maximum (FWHM) of ~10 nm which makes these crystals ideal for diode pumping. For the two pump wavelengths used in the laser experiments, the absorption cross-sections are reported in Table 3 . Note that at 1941 nm the absorption is very similar for E//Nm and Np so that in the case of (unpolarized) diode pumping the two polarizations absorb equally.
Fig. 2 Absorption cross-section of the 5I8 ↔ 5I7 transition of Ho for E//Nm and E//Np in KREW: a) KYW, b) KGdW, and c) KLuW.
Table 3. Absorption cross-section of Ho for different pump wavelengths and E//Nm and E//Np
The emission cross-section was obtained by the reciprocity method as mentioned earlier and used for the calculation of the gain cross-section with the following expression:
Here σg is the gain cross-section, σa is the absorption cross-section, σe is the emission cross-section and β is the inversion rate. Figure 3 shows the calculated emission cross-section only for E//Nm in the 1800-2200 nm range because this polarization shows the highest values. We calculated a maximum emission cross-section of ~2.65 × 10−20 cm2 at 2056 nm for Ho:KYW, ~2.70 × 10−20 cm2 at 2054 nm for Ho:KGdW and ~2.45 × 10−20 cm2 at 2059 nm for Ho:KLuW for E//Nm. For the other two polarizations, Ng and Np the emission cross-sections are smaller. From Fig. 3 (d, e, f), positive gain on the 5I7↔5I8 transition is achieved between 2020 and 2040 nm depending on the host up to 2100 nm. Two local gain maxima are located at 2056 and 2073 nm for Ho:KYW, 2054 and 2071 nm for Ho:KGdW, and 2059 and 2078 nm for Ho:KLuW. The chosen inversion rates from 0.2 to 0.3 are realistic because they describe the laser results presented later.Fig. 3 Absorption, emission and gain cross-sections of the 5I8↔5I7 transition of Ho in KREW crystals for E//Nm.
CW laser operation was evaluated by plotting the output power (Pout) as a function of the incident pump power (Pin) and the estimated slope efficiencies with respect to the incident pump power (see Table 4 ). Figures 4 (a, b, c) show the output-input characteristics of the Ho:KREW lasers under in-band pumping by the diode laser while Figs. 4 (d, e, f) show the same characteristics in the case of in-band pumping by the diode pumped Tm:KLuW laser.
Table 4. Summary of laser results for the two pump sources
Fig. 4 Output power vs. incident pump power of the Ho:KREW lasers (diode pump and Tm:KLuW as pump source) for different output coupling Toc. In all cases RC = 50 mm.
In the case of diode pumping, Toc = 3% gave slightly better output power for all the three hosts compared to other output couplings. The nonlinear behavior of the input-output dependence is related to the shift of the diode laser wavelength with the current because at higher pump powers longer wavelengths are better absorbed according to the optical absorption spectra (see Fig. 2). In this case the slope efficiencies were calculated from linear fits starting from Pin = 7 W (see Table 4). In the laser experiments with Tm:KLuW laser pumping the incident pump power was limited to ~2.3W in order to avoid any risk of damage in the crystal because of the small spot size. Also here, in most cases Toc = 3% gave the highest output, but for Ho:KLuW, Toc = 3% and 5% performed similarly. Further power scaling seems feasible for both pumping schemes because no roll-off effect for thermal reasons is seen and the only limitation is the pump power available.
Laser wavelength shifting towards shorter values was observed with the increase of the transmission of the output coupler as typical for three level laser systems (see Table 4). This effect was most pronounced for an output coupler with Toc = 20% (not included in Fig. 4), following the gain curves for each host shown in Figs. 3 d, e, f). The explanation for the existence of two distinct maxima in these curves is related to the effect of the crystal field (Stark) splitting of the energy levels (such a detailed spectroscopic analysis will be published elsewhere). Thus, for KYW and KLuW with Toc ranging from 1.5% to 5%, the emission is attributed to transitions from the first three lowest Stark levels of the 5I7 multiplet to the highest Stark level of the ground state, 5I8, while for Toc = 20%, the emission is attributed to a transition from the fifth Stark level of 5I7 to the same Stark level of the ground state. For KGdW with Toc ranging from 1.5% to 5%, the emission is attributed to the first two lowest Stark levels of the 5I7 multiplet to the highest Stark level of the ground state while for Toc = 20%, the emission is attributed to a transition from the third Stark level of 5I7 to the same Stark level of the ground state. Table 4 also includes the maximum output powers and the slope efficiencies achieved with the two pump sources.
We also studied the influence of the output coupler RC of the Ho:KLuW laser, using RC = 25, 50 and 75 mm and Toc = 1.5% in the case of diode pumping. The results shown in Fig. 5 indicate weakly pronounced effect of the mode matching with slope efficiency with respect to incident power in the range of 8.1 – 9.5%.
Fig. 5 Input-output characteristics of the diode-pumped 3 at.% Ho:KLuW laser with different RC and Toc = 1.5% (Inset: typical laser spectrum centered at 2080 nm).
To estimate the slope efficiencies with respect to the absorbed pump power only Toc = 3% is considered because this output coupling was optimum for both pump sources. For diode pump only single pass pumping is possible because the pump beam is very divergent and the non-absorbed part cannot be retro-reflected by the output coupler back to the crystal. However, in the case of Tm:KLuW laser pumping, the pump beam is retro-reflected and double pass absorption occurs because the output coupler reflects nearly 97% of the pump radiation. The measured single pass absorption of the crystals for the diode laser pump amounted to 14.8% for Ho:KYW, 15.7% for Ho:KLuW, and 14.9% for Ho:KGdW. These absorption values were measured without lasing and being constant for the whole range of pump powers. For the Tm:KLuW laser pump, the single pass absorption amounted to 22.9% for Ho:KYW, 21.5% for Ho:KLuW, and 23.2% for Ho:KGdW. Using these values the absorbed power was calculated and Fig. 4 is re-plotted as Fig. 6 only for Toc = 3%. In this case, the absorption values are given for pump levels below laser threshold. The absorption dropped 60-70% from the threshold to the maximum pump power available depending on the host. The obtained slopes are also listed in Table 5 together with other laser characteristics to respect the absorbed power.
Fig. 6 Estimated slope efficiencies for 3 at.% Ho:KREW crystals with respect to the absorbed power. a) diode pumping, b) Tm:KLuW pumping.
Table 5. Summary of estimated laser characteristics for the two pump sources (Toc = 3%)
From Fig. 6 and Table 5, it is inferred that all three hosts perform almost equally well. As can be seen, the best performance in the case of diode pumping is achieved with Ho:KYW while for Tm:KLuW laser pumping, Ho:KLuW is superior but the difference is not essential.
4. Conclusion
In conclusion, the laser properties of Ho:KYW, Ho:KGdW, and Ho:KLuW crystals under identical experimental conditions were compared using two different pump sources, a diode laser operating at 1941 nm and a Tm:KLuW laser operating at 1946 nm. The three hosts perform similarly, with maximum internal slope efficiency achieved as high as ~62%. Further power scaling to reach the Watt level is planned by optimization of the thickness of the active elements and the pump laser wavelength according to the optical absorption spectrum of Ho in KREW crystals and employing a more powerful laser diode.
Acknowledgments
This work was supported by the Spanish Government under projects MAT2008-06729-C02-02/NAN, TEC2010-21574-C02-02, PI09/90527 and the Catalan Authority under project 2009SGR235. J. J. Carvajal is supported by the Education and Science Ministry of Spain and European Social Fund under the Ramon y Cajal program, RYC2006 – 858. V. Jambunathan would like to acknowledge the Spanish Ministry of Education student mobility program, TME2009-00417. We also acknowledge support from the EC’s Seventh Framework program (LASERLAB-EUROPE, grant agreement nº 228334) and the German-Spanish bilateral program Acciones Integradas DE2009-0002, DAAD ID 50279160. This work has been partially funded by the European Commission under the Seventh Framework Programme, under project Cleanspace, FP7-SPACE-2010-1 –GA- 263044.
References and links
1. K. Scholle, S. Lamrini, P. Koopmann, and P. Fuhrberg, “2 μm laser sources and their possible applications,” Frontiers in Guided Wave Optics and Optoelectronics , ISBN 978–953–7619–82–4, pp. 471, February 2010.
2. V. Sudesh and K. Asai, “Spectroscopic and diode-pumped-laser properties of Tm,Ho:YLF; Tm,Ho:LuLF; and Tm,Ho:LuAG crystals: a comparative study,” J. Opt. Soc. Am. B 20(9), 1829–1837 (2003). [CrossRef]
3. E. Sani, A. Toncelli, M. Tonelli, N. Coluccelli, G. Galzerano, and P. Laporta, “Comparative analysis of Tm-Ho:KYF4 laser crystals,” Appl. Phys. B 81(6), 847–851 (2005). [CrossRef]
4. P. A. Budni, C. R. Ibach, S. D. Setzler, E. J. Gustafson, R. T. Castro, and E. P. Chicklis, “50-mJ, Q-switched, 2.09-μm holmium laser resonantly pumped by a diode-pumped 1.9-μm thulium laser,” Opt. Lett. 28(12), 1016–1018 (2003). [CrossRef] [PubMed]
5. B. Q. Yao, L. L. Zheng, R. L. Zhou, X. M. Duan, Y. J. Zhang, Y. L. Ju, Z. Wang, G. J. Zhao, and Q. Dong, “Holmium laser in-band pumped by a thulium laser in the same host of YAlO3,” Laser Phys. 18(12), 1501–1504 (2008). [CrossRef]
6. D. Y. Shen, A. Abdolvand, L. J. Cooper, and W. A. Clarkson, “Efficient Ho:YAG laser pumped by a cladding-pumped tunable Tm:silica-fibre laser,” Appl. Phys. B 79(5), 559–561 (2004). [CrossRef]
7. K. Scholle and P. Fuhrberg, “In-band pumping of high-power Ho:YAG lasers by laser diodes at 1.9 μm,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference, OSA Technical Digest (CD) (Optical Society of America, 2008), paper CTuAA1.
8. J. Liu, V. Petrov, X. Mateos, H. Zhang, and J. Wang, “Efficient high-power laser operation of Yb:KLu(WO4)2 crystals cut along the principal optical axes,” Opt. Lett. 32(14), 2016–2018 (2007). [CrossRef] [PubMed]
9. X. Mateos, V. Petrov, J. Liu, M. C. Pujol, U. Griebner, M. Aguiló, F. Díaz, M. Galan, and G. Viera, “Efficient 2 µm continuous-wave laser oscillation of Tm3+:KLu(WO4)2,” IEEE J. Quantum Electron. 42, 1008–1015 (2006). [CrossRef]
10. S. Rivier, X. Mateos, Ò. Silvestre, V. Petrov, U. Griebner, M. C. Pujol, M. Aguiló, F. Díaz, S. Vernay, and D. Rytz, “Thin-disk Yb:KLu(WO(4))(2) laser with single-pass pumping,” Opt. Lett. 33(7), 735–737 (2008). [CrossRef] [PubMed]
11. V. Petrov, M. C. Pujol, X. Mateos, Ó. Silvestre, S. Rivier, M. Aguiló, R. M. Solé, J. Liu, U. Griebner, and F. Díaz, “Growth and properties of KLu(WO4)2, and novel ytterbium and thulium lasers based on this monoclinic crystalline host,” Laser Photon. Rev. 1(2), 179–212 (2007). [CrossRef]
12. X. Mateos, V. Jambunathan, M. C. Pujol, J. J. Carvajal, F. Díaz, M. Aguiló, U. Griebner, and V. Petrov, “CW lasing of Ho in KLu(WO4)2 in-band pumped by a diode-pumped Tm:KLu(WO4)2 laser,” Opt. Express 18(20), 20793–20798 (2010). [CrossRef] [PubMed]
13. S. A. Payne, L. L. Chase, L. K. Smith, W. L. Kway, and W. F. Kruple, “Infrared cross-section measurements for crystals doped with Er3+, Tm3+, and Ho3+,” IEEE J. Quantum Electron. 28(11), 2619–2630 (1992). [CrossRef]
14. R. Solé, V. Nikolov, and X. Ruiz, “Growth of β-KGd1-xNdx(WO4)2 single crystals in K2W2O7 solvents,” J. Cryst. Growth 169(3), 600–603 (1996). [CrossRef]
