We report a spectroscopic investigation and an efficient Ho:LiLuF4 laser in-band pumped at 1938 nm. This represents the first laser emission of a fluoride crystal grown by micro–Pulling Down method in the 2 μm wavelength range. The Ho:LiLuF4 laser yielded a maximum output power of 7.1W with a slope efficiency of 41% and a threshold around 5W, at lasing wavelength of 2054.2 nm.
©2012 Optical Society of America
Mid-infrared (Mid-IR) solid-state lasers emitting in the eye-safe 2 μm spectral region have attracted more and more attention in recent years, because of their huge potential for applications in medicine, LIDAR systems, security, and high resolution spectroscopy for atmospheric pollution monitoring. At present the most utilized laser active media are based on fluoride or oxide crystals doped with trivalent rare earth ions. The main infrared rare earth emitters are Ho, Tm, Er, Dy, Nd and Yb whose laser channels are in the range of 1-4 µm. In the 2 µm region Ho and Tm are largely the most important dopants, and in literature the first laser emission at this wavelength goes back to 1962 with a Ho3+ doped CaWO4 crystal . Holmium lasers [2–4] are however preferred over thulium ones due to their longer operating laser wavelength, benefiting from higher atmospheric transmission and lower absorption in nonlinear crystals for Mid-IR light generation. The holmium 5I7 metastable level has a very long lifetime (typically around 14 ms) and a strong emission cross section, offering very good characteristics for generating high pulse energies when operating the laser in Q-switched mode [5–8]. Fluoride materials are particularly suitable to develop lasers operating in the Mid-IR because of their low phonon energy (300-500 cm−1) with respect to oxides (800-1000 cm−1) that decreases the detrimental effect of non radiative transition which could quench the upper laser level . Furthermore, the refraction index of fluorides decreases with temperature, leading to a negative thermal lens that is partly compensated by a positive lens effect due to end face bulging .
In this work we present the first, at the best of our knowledge, laser emission obtained from a Ho3+:LiLuF4 (Ho:LLF) sample grown by the micro Pulling Down (μ-PD) technique. To have a better outlook of the Ho:LLF sample quality, its optical properties will be compared with a sample grown by the well-established Czochralski (CZ) technique, which allows the growth of massive boules. µ-PD technique, developed by Fukuda , is instead very useful in developing fiber and rod shaped crystals [12–16]. The aim of this paper is to assess the possibility to grow laser grade fluoride materials with μ-PD technique. This aspect is important because it is possible to grow fibers with diameter as low as 500 μm, that cannot be obtained from massive boules, allowing a more effective cooling in high power applications. Moreover the µ-PD is cheap because a very small amount of raw material can be loaded in the crucible and it can be completely converted into single crystal. So μ-PD represents an appealing alternative technique to grow this, or other hosts, in developing laser materials.
The spectroscopic measurements have been performed in order to evaluate the potential of the Ho:LLF crystal as laser material in the infrared region. Because the Ho3+ doped materials are well known in literature, the detailed study was limited to the laser transition in order to compare the performances of our shaped crystal with those relative to a sample grown by a standard CZ technique. The absorption and emission spectra give us information about the optical quality of the sample and also about the fluorescence efficiency. The lifetime value allows us an estimation of the energy storage in the laser cavity and also to calculate the emission cross section.
2. Crystal growth
The samples under investigation were LLF single crystals doped with a concentration of 0.25 at. % of Ho3+ (3.6 · 1019 cm−3) and were grown in the Physics Department - National Enterprise for Nano Science and Technology laboratories. LLF is isomorphic to the well known LiYF4 which has the scheelite structure. The LLF lattice parameters are a = 5.167 Å and c = 10.735 Å and the symmetry group is I41/a; each elementary cell contains four formula units. The Ho3+ dopant substitutionally enters in the S4 symmetry Lu3+ sites having, in fluoride crystals, coordination number equal to 8. Since in this configuration the ionic radii of Lu3+ and Ho3+ in fluorides are close, 1.11 and 1.16 Å respectively , we assume a segregation coefficient equal to 1. The μ-PD technique however force the dopant to enter in the host matrix, so usually the segregation is always close to 1 even when the CZ method give lower coefficients.
The choice of fluoride materials means a special care to avoid contaminations, in particular the quality of the vacuum system (pressure limit below 10−7 mbar). Moreover the raw powders and Argon utilized for the growth processes have a purity of 5N (99.999%). This care is necessary to avoid the OH- contamination inside the crystal, which can affect irreparably the laser performance. The Ho:LLF was grown in a home made μ-PD furnace with RF heating in a glassy carbon crucible using LiF and LuF3 powders as raw materials for the crystal and the doping concentration was achieved adding the proper amount of HoF3 powders. The crucible shape was conical in the bottom, with an orifice of 1.3 mm diameter, and its volume was about 2.7 cm3.
We have performed two different growths with a pulling rate of about 0.5 mm/h, and the temperature of the melt was about 1120 K. The average size of the LLF shaped crystals, measured with a micrometer gauge having 0.01 mm resolution, was about 1.86 mm in diameter in both cases with 0.04 mm rms fluctuation. The lengths of the two fibers were ~40 and ~60 mm respectively. The single crystalline character of the samples was checked using a X-ray Laue technique that allows us to identify the crystallographic c-axis of the crystal that is tilted of about 6° with respect to the plane perpendicular to the rod axis. The grown crystals were of high optical quality, free of cracks and microbubbles. As it can be observed in Fig. 1 the samples show a rough surface. This feature is due to excess of LiF precipitated during the cooling of our samples .
However the inner part of the fiber is optically transparent and its optical quality has been checked measuring the distortions caused by the samples on the propagation of a TEM00 helium neon laser beam. The TEM00 mode was selected by using a pinhole, as described in , and the outcoming beam dimensions were regulated in order to obtain a spot slightly smaller than the sample diameter. The beam quality was measured with a Coherent Mode Master MMH-2S obtaining an M2x = 1.04 and M2y = 1.06 without the fiber. The measurements as been repeated inserting a 9 mm long fiber sample, cut from the same crystal utilized for the laser experiments. An M2x = 1.08 and M2y = 1.12 was obtained. The measurements show a small degradation of the beam quality confirming the absence of significant defects inside the sample. Moreover Fig. 2 (a) shows the intensity profile of the beam, utilizing a BeamScope-P8, while Fig. 2 (b) emphasizes the beam transmitted inserting a LLF sample. The beam profile confirms the absence of significant distortion introduced by the fiber on the laser beam propagation. From the crystals we cut two samples for laser test having dimensions diameter (D) = 1.8 mm, length (L) = 21 and 41 mm respectively. For spectroscopic measurements, a smaller sample (L = 5.75 mm), with the same orientations, has been prepared.
3.1 Absorption Spectrum
A spectroscopic analysis has been performed on 5I7 manifold in order to characterize the laser transition (5I7 → 5I8). Room temperature polarized absorption measurements were performed on the Ho:LLF sample by a CARY 500 spectrophotometer in the range 250 – 2200 nm in order to check the absorption bands of the Ho3+ and also to verify the absence of lines belonging to pollutants, that could be introduced inside the crystal during the growth and that can compromise the optical performances. In the following we do not report the results of this measurement, but within the sensitivity of the spectrophotometer, we recognized only the bands of the Holmium, a sign of the high purity of the sample. We recorded detailed spectra from 1800 to 2200 nm relative to the 5I8 → 5I7 transition.
We started the spectroscopic characterization of the sample measuring the polarized absorption spectra (E || c (π), E ⊥ c (σ)) at room temperature and they were acquired with a resolution of 0.6 nm. The absorption bands have been compared with those relative to a sample grown with the CZ technique, having the same doping level within the experimental errors. As it can be observed in Fig. 3 the spectra are comparable underlining the performances of the μ-PD sample. The slight difference in absorption between the two samples could be due to a small variation in the Ho3+ doping level.
3.2 Emission Spectrum
Fluorescence measurements have been performed at room temperature exciting the sample with a NICHIA blue diode laser tuned at 445 nm, according to the absorption spectrum of Ho:LLF. The exciting beam was focused on the sample by a 10 cm focal length lens. Luminescence was chopped and focused on the entrance slit of a monochromator equipped with a 300 lines/mm grating, blazed at 2 μm, and suitably filtered with an AR coated Ge filter. To record the spectra as a function of the orientation of the sample we used a Glan-Thomson polarizer in front of the input slit of monochromator. Our attention was focused on the transitions 5I7 → 5I8 in the range 1800-2200nm, using an InSb detector, cooled at liquid nitrogen temperature. The fluorescence signals were acquired by a SR830 Lock-in Amplifier and subsequently stored on a PC. The spectra were normalized for the optical response of the system using a black-body source at 3000 K.
We studied in detail the polarized fluorescence of the laser transitions 5I7 → 5I8 with a resolution of 1.5 nm. As an example, just to underline the same behaviour, we show in Fig. 4 the comparison, acquired in π−polarization, between the fluorescence yield of the μ-PD sample and the CZ sample, having the same doping level and in similar experimental conditions.
The two spectra have been normalized to the maximum at 2068 nm being our purpose just a comparison on peaks shape and position. We want to underline that the CZ sample comes from the crystal utilized in  where it demonstrates a very good laser performance. In Fig. 4 it is possible to observe that the two spectra are comparable, sign of a good quality of the μ-PD sample.
3.3 Fluorescence Lifetime
The fluorescence decay-time for 5I7 manifold has been measured at room temperature with the same experimental apparatus just described but the pump source was a pulsed tunable Ti:Al2O3 laser with a 10 Hz repetition rate and 30 ns pulse. In lifetime measurements the excitation wavelength was about 892 nm pumping the 5I5 manifold, and the pulse energy was reduced as much as possible to minimize power dependent effects. To reduce an artificial lifetime lengthening due to radiation trapping laser beam was focused near the edge of the crystal utilized for light collection. Moreover a second set of measurements was performed utilizing the pin-hole method  and the extrapolated lifetime is compared with the previous one. Measurements were performed utilizing four pin-hole width diameters 1.5, 1.0, 0.8 and 0.6 mm. To prevent detection of scattered laser light the collected radiation was suitably filtered in both experimental setup. Signal was detected by an InSb detector amplified with an overall response time of about 3μs. Data acquisition was performed with a digital oscilloscope connected to a computer.
The data were collected with the lowest possible excitation energy in order to avoid the nonlinear processes, as upconversion, and the decay curves shown a single exponential behaviour. The lifetime of the 5I7 manifold measured with the first method was 13.0 ± 0.8 ms, while that obtained with the pin-hole method was 13.1 ± 0.7 ms, in good agreement between them and in fair agreement with that measured for the CZ sample of 14.1 ± 0.8 ms .
4. Laser Experiments
4.1 Experimental Setup
Laser performance of both µ-PD grown crystals was tested in a ring laser setup shown in Fig. 5 . A single 80 W Tm-fiber laser (Model TLR-80-1940, from IPG Photonics) was used as pump source. The fiber laser operates in a wavelength range between 1937.6 nm at threshold and 1938 nm at maximum power of 85 W with a linewidth of 0.4 nm (FWHM) matching a small transparent window between the two water vapor absorption lines at 1937.5 nm and 1938.6 nm, while being well centered within the Ho:LLF absorption near 1940 nm. The unpolarized pump light was collimated and sent through a telescope consisting of two lenses with respective focal lengths of 200 mm and −50 mm. The spot radius of the pump beam was measured to be ~0.5 mm. End faces of both crystals were polished and no antireflection coating was used. The crystals were glued into a water cooled copper mount maintained at a temperature of 18°C and positioned between two flat dicroic mirrors (M1) with high reflectivity (R>99.8%) in the 2050 - 2100 nm wavelength range and high transmission (T>99.5%) at the pump wavelength. Both mirrors (M1) were tilted with approximately 10° to the optical axis. The transmitted pump light was measured behind mirror M1 during laser operation providing the calculation of laser performance as a function of the absorbed pump power, too. The physical length of the ring resonator was ~820 mm long with a convex high-reflector (M2) with a radius of curvature of 500 mm and a flat output coupler (OC), both were tilted with approximately 10° to the optical axis. Between mirror M2 and OC a Brewster-cut acousto optic modulator (AOM) was located which has not been used in the presented work but could provide Q-switching operation in future experiments. The crystal c-axis was set in the horizontal position (in the Brewster-plane of the AOM). Neglecting thermal lensing, the calculated TEM00 beam radius in the LLF crystal was 415 µm. A flat high reflector M3 has been used to force unidirectional oscillation in the ring cavity. For this configuration the laser was operated on π - polarization (polarization parallel to the crystalline c-axis).
Output couplers with a transmission of TOC = 1, 5, 9 and 20% were used. CW laser performance for the 21 mm long crystal is shown in Fig. 6 (a) as a function of total pump power incident on the crystal and (b) with respect to the absorbed pump power. Straight lines are result of a linear fit and the calculated slope efficiencies are given in the figures. At a maximum incident pump power of 85 W a maximum output power of 7.1 W was obtained with the TOC = 9% output coupler and a maximum slope efficiency of 41.4% was measured with respect to absorbed pump power. In Fig. 6 (a) the slope efficiency of TOC = 20% is lower that for TOC = 5%, while in Fig. 6 (b) the contrary occurs. This can be explained by the fact that higher output coupling increases the laser threshold and therefore the pump intensity is closer to the saturation intensity leading to higher bleaching of the laser crystal . Therefore, each curve in Fig. 6 (a) scales with a different absorption value. The laser output spectrum consists of several emission lines centered around 2067.9, 2067.6, 2054.2 and 2053.2 using the output couplers with TOC of 1, 5, 9 and 20%, respectively. No damage of the fluoride crystal has been observed in cw pumping up to the maximum pump power of 85 W. We estimated the maximum stress in the crystal by using LASCAD . With an absorbed pump power of 23 W and a fractional heat load of 6% due to the quantum defect, the theoretical maximum stress in the crystal is 2 MPa. This is well under the thermal fracture limit of 40 MPa given in .
By evaluating the slope efficiencies in the linear regime of the output couplers with TOC = 5% and 1%, one can fit the resonator losses as well as the intrinsic slope efficiency to the experimental points. The intrinsic slope efficiency is the limiting slope efficiency which can be achieved in the absence of resonator losses. To minimize the experimental error, the actual transmission of both output couplers has been measured directly with a power meter using the laser beam at wavelengths of 2067.6 and 2067.9 nm and tilting the output coupler with an angle of incidence of 10°. Plotting the inverse of the slope efficiency versus the inverse of the output coupler transmission, also known as a Caird plot , a value of 3.2% for the resonator losses and an intrinsic slope efficiency of 57.9% was found. It should be mentioned that the Caird plot analysis is only truly valid if the conditions of the output stay the same – e.g the lasing wavelength. However, the theoretic slope efficiency scales with the quotient of pump and laser wavelength  and therefore the small change in laser wavelength should be negligible. An explanation for the low resonator losses could be that both entrance faces of the crystal are nearly parallel and the crystal can act as an etalon (only one back reflection is observed using a beam of a HeNe laser). For comparison, laser experiments with a 30 mm long, 0.5 at. % Ho:LLF crystal made from bulk material has been done with the same resonator and pump configuration resulting in a value of 1.2% for the resonator losses and an intrinsic slope efficiency of 73.4% which is close to the performance presented in .
Brightness determination was done by measuring the beam diameter after it has been passed through a positive lens and by fitting the standard Gaussian beam propagation expression to the measured data shown in Fig. 7 . The beam quality has been measured at maximum laser power of 7.1 W and the M2 values found were M2x = 1.05 and M2y = 1.06.
The same experiments have been done using the 41 mm long Ho:LLF crystal. CW laser performance is shown in Fig. 8 (a) as a function of total pump power incident on the crystal and 8 (b) with respect to the absorbed pump power. Even though the crystal is longer and more pump power was absorbed the observed laser threshold was higher than with the 21 mm long crystal. The crystal had to be tilted a little bit to provide that the pump beam was centered both on entrance and exit face of this crystal (two back reflections have been observed using a beam of a HeNe laser resulting in an angle of 16 mrad between both faces). A maximum output power of 6.2 W was obtained with the TOC = 9% output coupler at maximum pump power of 85 W. A highest slope efficient of 27.7% was measured with respect to absorbed pump power with the TOC = 20% output coupler. The laser output spectra did not change significantly and the emission lines were centered around 2066.0, 2066.1, 2065.6 and 2065.6 nm using the output couplers with TOC of 1, 5, 9 and 20%, respectively. This indicates that the wedged crystal acts as a wavelength selective element in the ring cavity. A Caird analysis yielded a value of 8.9% for the resonator losses and an intrinsic slope efficiency of 58%. Due to the wedged faces of this crystal the resonator losses are considerable higher, however, the intrinsic slope efficiency of 58% is identical observed with the short crystal. Therefore the material properties of both crystals are the same. The worse results are associated to a deficient sample preparation of the long crystal.
At the moment it is not clear to understand the low intrinsic slope efficiencies observed with both Ho:LLF crystals grown by the μ-PD method. Excited state absorption can be excluded and at this low doping level energy-transfer upconversion should be negligible and the measured fluorescence lifetime of 13 ms is only slightly shorter with that measured for the CZ sample of 14.1 ms .
The beam quality has been measured at maximum laser power of 6.2 W and the M2 values found were M2x = 1.08 and M2y = 1.12, shown in Fig. 9 .
We achieved the first laser operation from Ho:LLF crystals grown by µ-PD technique. We applied a fiber laser pump setup using a ring laser cavity to characterize laser performance of two uncoated crystals with different length. A maximum output power of 7.1 W with a slope efficiency of 41.4% with respect to absorbed pump power was obtained at lasing wavelength of 2054.2 nm with a 21 mm long crystal. With the 41 mm long crystal more pump power is absorbed but the maximum output power was limited to 6.2 W due to higher resonator round trip losses. The intrinsic slope of ~58% was determined to be the same for both crystals and the beam quality was nearly diffraction limited (M2 < 1.12). These results emphasize the optical quality of the µ-PD grown crystals and higher output power should be reached using AR coating of the entrance and exit faces.
Authors wish to thank I. Grassini for preparing the samples and Dr. D. Parisi for helpful discussions.
References and links
1. L. J. Johnson, G. D. Boyd, and K. Nassau, “Optical maser characteristics of Ho3+ in CaWO4,” Proc. IRE, 50, 87 (1962).
2. B. M. Walsh, “Review of Tm and Ho Materials; Spectroscopy and Lasers,” Laser Phys. 19(4), 855–866 (2009). [CrossRef]
3. M. Eichhorn, “Quasi-three-level solid-state lasers in the near and mid infrared based on trivalent rare earth ions,” Appl. Phys. B 93(2-3), 269–316 (2008). [CrossRef]
4. N. P. Barnes, B. M. Walsh, and E. D. Filer, “Ho:Ho upconversion: applications to Ho lasers,” J. Opt. Soc. Am. B 20(6), 1212 (2003). [CrossRef]
5. 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]
6. A. Dergachev, D. Armstrong, A. Smith, T. Drake, and M. Dubois, “3.4-μm ZGP RISTRA nanosecond optical parametric oscillator pumped by a 2.05-μm Ho:YLF MOPA system,” Opt. Express 15(22), 14404 (2007). [CrossRef]
9. F. Cornacchia, A. Toncelli, and M. Tonelli, “2-μm lasers with fluoride crystals: Research and development,” Prog. Quantum Electron. 33(2-4), 61–109 (2009). [CrossRef]
10. R. L. Aggarwal, D. J. Ripin, J. R. Ochoa, and T. Y. Fan, “Measurement of thermo-optic properties of Y3Al5O12, Lu3Al5O12, YAlO3, LiYF4, LiLuF4, BaY2F8, KGd(WO4)2, and KY(WO4)2 laser crystals in the 80–300 K temperature range,” J. Appl. Phys. 98, 103514 (2005). [CrossRef]
11. D. H. Yoon, I. Yonenaga, T. Fukuda, and N. Ohnishi, “Crystal growth of dislocation-free LiNbO3 single crystals by micro pulling down method,” J. Cryst. Growth 142(3-4), 339–343 (1994). [CrossRef]
12. D. Sangla, I. Martial, N. Aubry, J. Didierjean, D. Perrodin, F. Balembois, K. Lebbou, A. Brenier, P. Georges, O. Tillement, and J. M. Fourmigue, “High power laser operation with crystal fibers,” Appl. Phys. B 97(2), 263–273 (2009). [CrossRef]
13. D. Maier, D. Rhede, R. Bertram, D. Klimm, and R. Fornari, “Dopant segregations in oxide single-crystal fibers grown by the micro-pulling-down method,” Opt. Mater. 30(1), 11–14 (2007). [CrossRef]
14. K. Y. Huang, K. Y. Hsu, D. Y. Jheng, W. J. Zhuo, P. Y. Chen, P. S. Yeh, and S. L. Huang, “Low-loss propagation in Cr4+:YAG double-clad crystal fiber fabricated by sapphire tube assisted CDLHPG technique,” Opt. Express 16(16), 12264–12271 (2008). [CrossRef] [PubMed]
15. A. Yoshikawa and V. Chani, “Growth of Optical Crystals by the Micro-Pulling-Down Method,” MRS Bull. 34(04), 266–270 (2009). [CrossRef]
16. M. H. Pham, M. M. Cadatal, T. Tatsumi, A. Saiki, Y. Furukawa, T. Nakazato, E. Estacio, N. Sarukura, T. Suyama, K. Fukuda, K. J. Kim, A. Yoshikawa, and F. Saito, “Laser quality Ce3+:LiCaAlF6 grown by micro-pulling-down method,” Jpn. J. Appl. Phys. 47(7), 5605–5607 (2008). [CrossRef]
17. A. A. Kaminskii, “Laser crystals,” Springer-Verlag Series in Optical Science v.14 (1990).
18. P. Rogin and J. Hulliger, “Liquid phase epitaxy of LiYF4,” J. Cryst. Growth 179(3-4), 551–558 (1997). [CrossRef]
19. S. Veronesi, Y. Z. Zhang, M. Tonelli, A. Agnesi, A. Greborio, F. Pirzio, and G. Reali, “Spectroscopy and efficient laser emission of Yb3+: LuAG single crystal grown by μ-PD,” Opt. Commun. 285(3), 315–321 (2012). [CrossRef]
21. H. Kühn, S. T. Fredrich-Thornton, C. Kränkel, R. Peters, and K. Petermann, “Model for the calculation of radiation trapping and description of the pinhole method,” Opt. Lett. 32(13), 1908–1910 (2007). [CrossRef] [PubMed]
22. J. W. Kim, J. I. Mackenzie, D. Parisi, S. Veronesi, M. Tonelli, and W. A. Clarkson, “Efficient fiber-laser pumped Ho:LuLiF4 laser,” Solid State Lasers and Amplifiers IV, and High-Power Lasers. Edited by Graf, Thomas; MacKenzie, Jacob I.; Jelinková, Helena; Paulus, Gerhard G.; Bagnoud, Vincent; Le Blanc, Catherine. Proceedings of the SPIE, Volume 7721, pp. 77210V (2010).
23. M. Schellhorn, “A comparison of resonantly pumped Ho:YLF and Ho:LLF lasers in CW and Q-switched operation under identical pump conditions,” Appl. Phys. B 103(4), 777–788 (2011). [CrossRef]
24. LASCAD, LAS-CAD GmbH, Brunhildenstrasse 9, 80639 Munich, Germany, http://www.las-cad.com.
26. J. A. Caird, S. A. Payne, P. R. Staber, A. J. Ramponi, L. L. Chase, and W. F. Krupke, “Quantum electronic properties of the Na3Ga2Li3F12:Cr3+ laser,” IEEE J. Quantum Electron. 24(6), 1077–1099 (1988). [CrossRef]