Open Access
Add to BibSonomyAbstract
We report a rigorous study of the spectroscopic, laser and thermal properties of a 10at.% and a 15at.% Yb:LuAG crystals. A loss mechanism is observed in the medium with the highest doping, pumped at 936 nm and 968 nm, as a sharp and dramatic decrease of the laser output power is measured at higher excitation densities. The nonlinearity of the loss mechanism is confirmed by the fluorescence data and by the thermal lens. In particular, the dioptric power of the thermal lens acquired at different pumping levels shows a strong deviation of the expected linear trend. Here we report the influence of both the concentration and the ion excitation density of Yb3+ on the output powers, the slope efficiencies and the thresholds. Conversely excellent results are achieved with the 10at.%, which does not show any loss mechanism as at 1046 nm it delivers 11.8 W with a slope efficiency of ηs = 82%, which is, to the best of our knowledge, the highest value reported in literature for this material.
© 2014 Optical Society of America
1. Introduction
One of the most important requirements to develop high performance ytterbium thin-disk lasers is the use of thin gain media with high concentration of dopant. Thinner samples permit a better management of the heat load while the high level of doping allows the crystal to achieve higher absorption levels. For this purpose several hosts able to support high Yb doping level were proposed in literature with different compositions [1], such as garnets [2,3], fluorides [4,5], tungstates [6]. However, high doping levels mean shorter distances between the ytterbium ions with a consequent modification of the interactions between ion-ion and ion-host, which influence the matrix properties by reducing, for instance, its thermal conductivity [7]. Reduced heat dissipation leads to a temperature increase in the pumped region, which, in turn, reduces the absorption cross section and the emission cross section at the pump and laser wavelengths, respectively [8]. Furthermore, dangerous thermal lensing effects can occur.
Beside this, in heavily Yb-doped oxides an additional nonlinear loss mechanism has been discovered. In 2005 Larionov et al. [9] observed a non-linear decays of the excited states of Yb3+-ions in two crystals of Yb:YAG doped with Yb 12.7at.%, and 15.7at.%, respectively, by measuring the gain and the temperature of the sample surfaces. The rate of these processes was found to be dependent on the doping concentration, on the density of the excited ions and on the temperature. The role played in the phenomenon by Yb3+ was observed by Ueda et al. [10] testing two highly-doped ceramics, i.e. Yb(10at.%):Lu2O3 and Yb(15at.%, 20at.%):Y2O3. Bisson et al. [11] using the same matrices found a switch of the thermal emission and the electrical conductivity only when pumping on the absorption band of the dopant. Strong photocurrents were measured also in Yb(10 at.%):Lu2O3 and Yb:YAG with several doping levels up to 40at.% by Brandt et al. [12].
The first observation of a nonlinear loss-mechanism during the laser action was done in 2010. By using Yb(20at.%):YAG ceramic Pirri et al. [13] observed a sudden decrease of the laser efficiency at high ion excitation density (ρexc = 4.5x1020 cm−3). The experimental results, obtained pumping the ceramic at 940 nm, confirmed the fundamental role of Yb3+ ions, and moreover, the dependence of the phenomenon on the levels of dopant, on the ion excitation densities and on the temperature experienced by the sample. In particular, it is demonstrated that the role played by each of these parameters is different as only the first two trigger the loss mechanism.
The data seem to suggest the common origin of a mechanism in high Yb-doped oxides matrices underpinning both the loss of the laser efficiency and the appearance of the photocurrent.. In a deeper analysis there is not any experimental proof supporting this common origin because the photocurrent is measured in samples placed outside of the cavity. Conversely, the nature of the host, crystals or ceramic, does not influence this loss mechanism.
This article is stimulated by two different needs. The first is to investigate the performance of the Yb:LuAG with concentrations of 10at.% and 15at.% as LuAG shows similar physical properties to the most widely used YAG and higher thermal conductivity, totally independent on the doping level [14-18].We compared the performance of these two crystals, obtained in the same experimental conditions, in term of maximum output powers, slope efficiencies, laser threshold.
The second is to study the appearance of the mentioned loss mechanism in the heavily doped Yb (15at.%):LuAG crystal. We investigated this phenomenon by pumping the sample at 936 nm as well as on the main absorption peak placed at 968 nm, in quasi-CW and CW pumping regime.
Finally, we measured the dioptric power of the thermal lens in 10at.%, 15at.% as well as in 20at.% samples, which allows to investigate the thermal dissipation mechanisms in dependence on the concentration of Yb ions.
2. Spectroscopic investigation of 10at.% and 15at.% crystals
The crystals were grown by Crytur spol. s r.o. (Czech Republic) using the Czochralski method, in an iridium crucibles, under the atmosphere of argon with 1 - 1.5% of oxygen. Raw materials (Lu2O3, Al2O3 and Yb2O3 powder oxides) were of 5N purity, but not necessarily from the same batch. The cut optical elements from the parent crystal boule were annealed under air (temperature 1300 - 1400 °C) for typically 24 hours to completely oxidize Yb to Yb3+ charge state and remove oxygen vacancies. Annealed elements were mechanically polished to the laser quality (at least λ/10).
Optical absorption and luminescence spectra in the UV/visible spectral region were measured to characterize the Yb3+ centers, the optical quality of the samples, and to find out other eventual luminescent impurities.
Figure 1 shows the absorption spectrum recorded with a Shimadzu spectrometer 3101PC at room temperature. The structured 4f-4f absorption lines of Yb3+ in near infrared region related to transitions within 2F7/2–2F5/2 multiples include an intense band within 915-945 nm, suitable for diode-laser pumping, and a well-defined peak at 968 nm, ascribed to the so-called zero-phonon absorption line, with a bandwidth of 3 nm.
Fig. 1 Absorption spectra of Yb-doped LuAG single crystals with concentration doping of 10at.% and 15at.%, in the interval 200-800 nm (a) and 800-1100 nm (b).
In near ultraviolet region, the edge of the Yb3+ Charge Transfer Transition (CTT) can be seen, as well as the absorption shoulder at about 262 nm in the 10at.% sample, ascribed to the Tb3+ impurity (4f–5d LS transition) following [19]. Smoothly increasing absorption in 15at.% crystal comes most probably from slightly inferior polishing quality.
Luminescence spectra were measured with a custom made 5000M spectrofluorometer from Horiba Jobin Yvon. In Fig. 2 the PhotoLuminescence Excitation (PLE) and emission (PL) spectra related to the Yb3+ CTT are provided. They are fully consistent with the existing literature data [20].
Fig. 2 Excitation (PLE) and emission (PL) spectra related to the CTT of Yb3+ center in 10at.% (a) and in 15at.% (b) crystals.
Figure 3 reports the PL and PLE spectra of 10at.%. A small contamination of Tb3+ is found. The peak in the PLE spectrum due to the 4f–5d LS transition coincides with the absorption shoulder in Fig. 1, and the peaks between 370 nm and 450 nm in the PL spectrum due to the transition from the 5D3 level to 7Fx ground state multiplet [21]. Tb3+ accidental impurity comes certainly from the raw Lu2O3 and Yb2O3 powders. Conversely, the 15at.% sample does not show any contamination due to terbium or other impurities. This difference between the samples in terms of Terbium content can be addressed to different batches of raw materials. Furthermore, segregation coefficient of Tb in LuAG host is less then 1 so that its content changes along the crystal boule. Consequently, Tb content may differ in elements taken from different relative positions in the crystal boule (this information is not available from the producer).
3. Laser investigation
In the first series of measurements we tested and compared the laser behaviour of the media pumped at 936 nm in quasi-continuous wave (QCW). For each measurement we collected the output power (Pout) at different levels of the absorbed pump power (Pabs) and the fluorescence light emitted perpendicularly to the propagation axis of the cavity. This latter was collected when the samples were either lasing or not lasing (herein named Laser ON/OFF). The measurements are done by using a set of OCs with various transmissions (from T~2% to T~86%) in order to find the best output coupling and, at the same time, to investigate the influence of the ion excitation density (ρexc) on the laser performance. In this respect, we remind that using OCs with higher transmissions the population inversion density ρexc needed to sustain the laser action increases. The crystal absorption at the pump wavelength is determined by measuring the residual pump radiation emerging from the FM. The slope efficiencies (ηs) are estimated by taking into account almost all points of the curves. We note that although the thickness of the crystals is different the use of the Pabs in the graphics allows a fair comparison among their laser behaviors.
In the second set of measurements we investigated the 15at.% crystal pumped on the zero-absorption line corresponding to 968 nm in CW and in QCW (DF from 20% to 40%, 10Hz). The experiment is performed by using the same experimental setup and by carefully following the same protocol of investigation used at 936 nm.
3.1 Experimental set-up
The experimental setup used to test the laser performance of the crystals is sketched in Fig. 4.
Fig. 4 Experimental set up to test the laser performance of the samples. EM: End Mirror (flat); FM: Folding Mirror; OC: Output Coupler (flat); M1, M2: power meters; F1,2 filters for rejection of pump radiation (b); PD1,2: photodiodes.
The resonator is constituted by three mirrors placed in a V-shaped configuration. EM is a flat mirror with a dichroic coating with high transmission around the pump wavelengths and high reflectivity in the laser emission band. The folding mirror, FM, has a curvature radius of 100 mm with high transmission at the pump wavelengths. OC is the output coupler mirror. The uncoated 10at.% and 15at.% Yb:LuAG crystals have a thickness of 2 mm and 1 mm respectively, and a side size of 5x5 mm2. They are soldered with indium on a copper heat sink which in turn is water cooled at 18 °C. We did not find significant variations in the laser performances when changing the temperature few degrees above or below this value. The crystals are placed as near as possible to the flat EM. A very careful orientation of the facets perpendicular to the cavity axis allowed the re-injection of the Fresnel reflection, which are estimated around of 8.5%.
For the pumping of these samples, two fiber-coupled laser diodes are used. The emission from the fiber is refocused and injected into the sample by a pair of achromatic doublets with a magnification of 1:1. The first emits at 936 nm (fiber core diameter 200 μm, numerical aperture of N.A. = 0.22, maximum pump power Pp = 25 W) and it has an almost Gaussian intensity distribution in the focal plane of the doublets with 150 μm of radius at 1/e2. The second emits at 968 nm (fiber core diameter 100 μm, N.A. = 0.15, Pp = 25 W); in the focal plane the pump beam has a fairly Gaussian distribution intensity with a spot radius around 67 μm at 1/e2. Both pump lasers can operate either in CW or quasi-CW regime. In the experiment we employed Duty Factor, DF, of 20% and 40%.
3.2 Laser performance at 936 nm
Figures 5(a) and 5(b) report the experimental results obtained with four different OCs (from T = 1.9% to T = 18.8%).
Fig. 5 Laser output power measured at low excitation density.(a): 10at.% doped sample, (b): 15at.% doped sample. T: output coupler transmission; λL: laser wavelength; ηs: slope efficiency. The crystals are pumped at 936 nm in quasi-CW (DF = 20%, 10 Hz). The unsaturated absorptions are 88.5% and 77.4% for the 10at.% and 15at.%, respectively.
Excellent performances are achieved by the 10at.% both in terms of slope efficiency and output powers. Emitting at 1046 nm the laser delivers Pout = 11.8 W with a slope efficiency of ηs = 82% while the threshold is Pth,abs = 0.5 W. The 15at.% crystal releases the maximum output power at 1030 nm, with Pout = 9.7 W. The slope efficiency is ηs = 59% while the threshold is Pth,abs = 0.57 W. The emission from both crystals shows a weak dependence on the OC transmission. In particular, with 10at.% the lowest slope efficiency and the minimum output power are ηs = 71% and Pout = 9.7 W (T = 1.9%, λL = 1049 nm), respectively. A decrease of 6% in the slope efficiency and of about 3 W in the output power are found with the 15at.% sample.
Figure 6 reports the results obtained involving higher Yb-ion excitation densities in the laser action, i.e. using output couplers with T = 86% and T = 57.7%. It can be seen that the output from the 10at.% doped sample linearly increases with the absorbed pump power, without any significant rollover. On the other hand, with the 15at.% doped sample two different regimes can be easily observed. At lower absorption levels, i.e. Pabs<13 W by T = 57.7% and Pabs<7.5 W by T = 86%, the increase of the temperature inside of the crystal does not affect the laser performance as the output power linearly increases with the pump power. The scenario totally changes as the Pabs exceeds the mentioned threshold with a dramatic decrease of the Pout.
Fig. 6 Laser output power versus absorbed pump power at high excitation density. Both crystals emit at 1030 nm. The crystals are pumped at 936 nm in quasi-CW (DF = 20%, 10 Hz).
We evaluated the excited Yb ions density occurring at the laser threshold with a rate equation model, obtaining ρexc = 4.45x1020 exc.ions/cm3 (15at.%, T = 86%) and ρexc = 2.89x1020 exc.ions/cm3 (10at.%, T = 57.7%). We note that, in absence of further loss mechanisms, the excitation density above threshold should remain constant at different pumping levels. The experimental data clearly show the dissimilar behavior of the crystals as a sharp decrease of the Pout is only observed in the heavily doped crystal.
Closing the cavity with an OC having T = 57.7%, we studied the dependence of this effect on the temperature experienced by the 15at.% sample, by increasing the DF from 20% to 40%, see Figs. 7(a) and 7(b). It is worth to note that, at the maximum level of the absorption used in the experiment (Pabs~16 W), the laser cavity still emits 2 W when the crystal is pumped with DF = 20%, whereas the laser action is practically hampered with DF = 40%. These two different regimes can be discriminated by looking at the fluorescence emission as well. Figure 7(b) reports the data acquired with DF = 40%. At low values of Pabs both curves increase linearly, until they reach a plateau, at a level of Pabs corresponding to the rollover in the laser emission. A similar behavior of the crystal is observed with DF = 20%.
Fig. 7 (a): Laser output power versus absorbed pump power at higher ion excitation density with two Duty Factor (20% and 40%); (b): the corresponding fluorescence light acquired when the crystal is lasing or switched off with DF = 40%.
3.3 Laser performance with 968 nm pump
Figures 8(a) and 8(b) reports the results obtained with 15at.% at 968 nm in QCW (DF = 20%, 10Hz) and CW respectively. As in the first campaign of measurements, at low excitation density the output power increases linearly with the pump power levels. At 1031 nm we measured Pout = 8 W with a corresponding slope efficiency of ηs = 61%.
Fig. 8 Laser output power obtained by several OCs with low transmission at DF = 20% (a) and in CW (b); (c) shows the laser output obtained with three different DF, with an OC with high transmission (i.e. at higher ion excitation density). The pump wavelength is λp = 968 nm. The absorption at the pump wavelength is 73%.
The threshold is Pabs,th = 0.9 W. In terms of maximum output power, we obtained a lower value with respect to the results obtained when pumping at 936 nm. This can be explained considering that the overall pump absorption of the crystal at 968 nm (73%) is lower than at 936 nm (77.4%), because the spectral width of the absorption peak at 968 nm is narrower (FWHM = 3 nm) than the width of the pump radiation (FWHM = 6 nm). For the CW emission at 1048 nm we measured 3 W (Pabs = 8.6 W) with a slope of ηs = 35%.
The laser output power as a function of the absorbed pump power using an output coupler with T = 57.7% is reported in Fig. 8(c). The measurements where done by employing three different DFs. Similarly with data obtained at 936 nm, at higher levels of absorbed pump power the laser efficiency decreases. The roll-over onset of the curve is shifted from Pabs~7 W with DF = 20% to Pabs~3 W with DF = 40%.
4. Thermal lens measurements
To shed light on the thermal behaviour of Yb:LuAG crystal, we measured the dioptric power of the thermal lens. We investigated samples with three different levels of doping, i.e. 10at.%, 15at.% and 20at.% (uncoated, 2 mm-length, 5x5 mm2 surfaces). The knowledge of the thermal lens, apart from its intrinsic scientific value, allows to estimate the radiative quantum efficiency and its dependence on the Yb concentration.
4.1 Experimental set-up
The measurements of the thermal lens were carried out in a different apparatus as reported in Fig. 9. The pump source is a fiber coupled semiconductor laser emitting at 936 nm. The output from the fiber (200 μm diameter, 0.22 NA) is refocused and magnified by a pair of achromatic doublets on a waist with 500 μm diameter at 1/e2 located in the center of the sample. DM is a dichroic mirror which steers the pump beam emerging from the refocusing optics in the laser cavity. The overall distance between the focusing mirror M2 and the end mirror EM is 113 mm, the distance between M2 and the output coupler OC is 210 mm, and curvature radius of M2 is 200 mm. This result in a fundamental mode cavity radius of 100 μm for the cold cavity, on the waist located at EM. The crystal is mounted on a heat sink as described in section 3.1.
The thermal lens is probed by a beam emitted by a HeNe laser. The beam is expanded and collimated, sent into the sample collinearly with the pump beam through the mirrors DM and EM, and extracted from the cavity through the mirror M1. The probe beam illuminates the whole clear aperture of the sample holder. The wavefront distortion induced by thermal effects is measured by means of a Shack-Hartmann wavefront sensor (ML4010 by Metrolux GmbH, area 8.98 x 6.71 mm2 and 200 μm pitch of the microlenses array).
In order to measure the wavefront distortion occurring over an area of about 500 μm diameter, the probe beam was expanded by an afocal telescope formed by the two positive lenses with nominal focal lengths f1 = 155 mm and f2 = 750 mm placed at a distance d = f1 + f2, see Fig. 9. The sample and the sensor are placed in the conjugate planes of the telescope. The probe beam at the output of the sample is then reimaged on the sensor with a magnification M = f2/f1, without any further wavefront curvature introduced by the telescope. The magnification of the telescope was accurately measured and it resulted M = 4.81. Several filters (not shown in Fig. 9 for clarity) are used to reject the residual pump beam that would otherwise impinge on the sensor. The thermal deformation of the wavefront is measured at several levels of incident pump power. A circular portion of the wavefront with radius a corresponding to the pumped area of the sample is then selected and analyzed by fitting with Zernike polynomials up to the 6th order. The effective dioptric power Dth of the thermal lens is then calculated with the formula:
where Z2,0 is the coefficient of the Zernike polynomial with radial index 2 and azimuthal index 0, corresponding to the wavefront defocus. In this analysis, we adopted a value of a = 1.2 mm (on the sensor plane), corresponding to the pump beam radius at 1/e2 (i.e. 250 μm) on the sample plane. Due to the different absorption featured by the two crystals (2 mm and 1 mm of thickness), the dioptric power of the thermal lens was analyzed as a function of the absorbed pump power.It must be noticed that in this configuration the pumping conditions are rather different from those used in the measurement described before, because they were optimized for the detection of the thermal lens rather than for the laser performance. In particular, for a given pump power level the pump power density is much lower (due to the larger pump beam cross section area) so that the pump power threshold for the laser action is higher. Furthermore, the different cavity geometry determines a larger fundamental mode radius on the sample than in the previous set up. Finally, the pumping regime is now truly CW (and no longer QCW), so that the overall average thermal load on the samples is in general higher than in the previous measurements. Due to these modifications, the laser output from the samples is rather different from the previous results, as it will be shown in the following parts.
4.2 Thermal lens measurements
Figure 10 reports the thermal lens measurements. Concerning the 10at.% sample, see Fig. 10(a), the lasing condition was obtained by an OC with low transmission (T = 2%) in order to maximize the intracavity circulating power and, in turn, the difference in thermal load between lasing and non-lasing conditions. In non lasing conditions the dioptric power D of the thermal lens is expected to scale linearly with the absorbed pump power Pabs, according to the following relation [22]:
Fig. 10 Thermal lens dioptric power as a function of the absorbed pump power for the 10at.% doped sample (a), 15at.% doped sample (b) and 20at.% doped sample (c). The solid curves are the second order polynomials or linear best fits.
The quantity in brackets expresses the fraction of the absorbed pump power which is converted in heat and it is influenced by the radiative quantum efficiency ηr; λP and λF are the pump wavelength and the baricenter wavelength of the fluorescence spectrum respectively. The constant Aoff incorporates the thermo-optical and the thermal properties of the material, as well as the geometric conditions [22].
In lasing conditions the overall probability of nonradiative decays is reduced by the stimulated emission probability that becomes dominant when the laser is pumped well above the threshold. In this latter condition the expression for the thermal lens can be modified as it follows:
where λL is the lasing wavelength.The thermal lens in the 10at.% doped sample shows the expected linear dependence from the absorbed pump power, see Fig. 10(a); the slope in lasing and nonlasing conditions is slightly different for the reasons expressed above. Conversely, the data obtained with the 15at.% and 20at.% crystals, see Figs. 10(b) and 10(c), show a dramatic departure from the expected linear trend, which increases with the Yb concentration. For these two samples it was possible to measure the thermal lens in nonlasing conditions only, as it was impossible to obtain laser action from the samples in these experimental conditions. The inhibition of the laser action (which was not observed with the experimental set up described in Section 3) must be probably ascribed to the different excitation conditions used in this set up, as discussed above: due to the lower pump power density the laser threshold is higher than in the previous set up. Moreover due to the higher thermal load determined by the CW pumping conditions, the resonator is more readily driven out of the stability region by the strong thermal lens.
5. Discussion
The remarkable results obtained with the 10at.% crystal such as high efficiencies (ηs = 82%), high output powers (Pout = 11.7 W) and low laser thresholds (Pth.abs< 0.6 W), confirmed the potentiality of the Yb:LuAG to be an effective candidate in all laser systems based on Yb. However, also in this matrix, high levels of doping can constitute a serious problem for lasing as additional nonradiative channels can appear with a consequent deterioration of the laser performance.
All the results obtained with the 15at.% crystal clearly demonstrate that a nonlinear loss mechanism takes place when higher ion excitation density are involved in the laser action. We measured a sudden decrease of the laser output power at higher levels of absorption, which is independent on the Yb-resonant pump wavelength (936 nm or 968 nm). The observed shift of the onset of the roll-over in the curves toward lower value of the absorption levels (from Pabs~13 W with T = 57.7% to Pabs~8W by T = 86%) demonstrates that it starts at a well-defined threshold of the excitation levels and its rate is enhanced by increasing ρexc. According to our calculation based on the equation rates, the estimated threshold of the density is around 4.5x1020exc. ion/cm3.
The presence of non-radiative channels is also supported by measurements of the collected fluorescence light. To focus on the measurements reported in Fig. 7(b), the first part of the curves can be easily understood as the increase of the pump power forces the Yb-ions to occupy the energetic states placed in the upper folder (2F5/2) with the subsequent de-excitation of the population. Conversely, the presence of the plateau indicates a loss of excited Yb3+ by some non-radiative mechanism.
In order to estimate the relative weight of this additional decay, we have estimated the radiative quantum efficiency (ηr) by using the thermal lens measurements. In particular, by the Eqs. (2) and (3), ηr can be calculated as:
where Δon and Δoff represents the slope of the curves reported in Fig. 10(a) (see also Eqs. (2) and (3)). The value of the constant Aon is not necessarily the same as Aoff, because the heat sources distribution in lasing conditions can be different from nonlasing conditions. Nevertheless if the laser is well above threshold we can make the approximation that the whole pumped region is lasing, so that the heat sources distribution becomes the same in the two cases and Aon = Aoff.In the case of 10at.% doping we calculated a radiative quantum efficiency of ηr = 0.96 with Δon = 1.12W−1m−1 and Δoff = 1.20W−1m−1, which is very close to the unity; the non-radiative decay processes plays then only a marginal role.
To focus on the 15at.% and 20at.% crystals, it has to be noticed that the nonlinear behaviour of the thermal lens with respect of the absorbed pump power indicates that the dissipated power fraction dramatically increases for increasing pump power, thus indicating a decrease of the radiative quantum efficiency. By inverting Eq. (2) the value of the radiative quantum efficiency ηr for each absorbed power level can be calculated as:
The value of Aoff can be derived from the measurements carried out on the 10at.% doped sample, because the geometrical conditions in our experiment remain the same, as well as the thermal conductivity of the sample (which in LuAG is almost independent from doping [15]).The results of these calculations are reported in Fig. 11. Both crystals show an increase of the thermal dissipation that, in turn, indicates a reduction in the fluorescence quantum efficiency for increasing absorbed pump powers. The decrease of ηr is significant in the case of the 15at.% doped sample (from about 0.9 to about 0.6 at 8.5 W of absorbed power) and it becomes dramatic for the 20at.% sample, which shows a decrease down to about 0.3 in the absorbed power range used in the experiments. We underline that even at low pumping levels the fluorescence quantum efficiency is lower than the value recorded with the 10at.% crystal.
Fig. 11 Fluorescence quantum efficiency as calculated from the thermal lens measurements for the 15at.% and 20at.% samples .
In principle, the superlinear dependence of the thermal lens from the absorbed power could be due to other reasons, such as to a dependence from temperature of the thermo-optical coefficients (increasing with temperature), or a reduction of the thermal conductivity with temperature. Nonetheless, these reasons can be ruled out on the basis of the temperature dependence of these parameters in LuAG [15].
6. Conclusions
The results obtained at low excitation density with both crystals confirm the full potential of the Yb:LuAG to be employed in the development of diode-pumped solid state laser sources. In particular, we measured with the 10at.% doping an output power of 11.8 W with a slope efficiency as high as ηs = 82% at 1046 nm. However, a heavy doping of the matrix can trigger loss mechanisms which degrade the laser performance. The presence of such nonradiative channels is supported by the thermal lens and fluorescence measurements. Although, at the present state of knowledge, there is not any theoretical model [12] explaining the origin of this mechanism, the occurrence of this deterioration of the laser performance in high purity samples (as pointed out from the spectroscopic measurements) suggests that this mechanism can hardly be related to some energy transfer mechanism involving impurities contained in the host [23].
Acknowledgments
The research was supported by Regione Toscana, project “CTOTUS-Progetto integrato per lo sviluppo della Capacità Tecnologica e Operativa della Toscana per l’Utilizzo dello Spazio” (POR FESR 2007-2013 Attività 1.1 Linea d'intervento D); by the Consiglio Nazionale delle Ricerche, CNR-RSTL “Ricerca Spontanea a Tema libero”, id. 959; by the joint project of ASCR and CNR and Czech GA AV project M100100910, 2012-2015.
References and links
1. T. Südmeyer, C. Kränkel, C. R. E. Baer, O. H. Heckl, C. J. Saraceno, M. Golling, R. Peters, K. Petermann, G. Huber, and U. Keller, “High-power ultrafast thin disk laser oscillators and their potential for sub-100-femtosecond pulse generation,” Appl. Phys. B 97(2), 281–295 (2009). [CrossRef]
2. U. Brauch, A. Giesen, M. Karszewski, C. Stewen, and A. Voss, “Multiwatt diode-pumped Yb:YAG thin disk laser continuously tunable between 1018 and 1053 nm,” Opt. Lett. 20(7), 713–715 (1995). [CrossRef] [PubMed]
3. J. Aus der Au, G. J. Spühler, T. Südmeyer, R. Paschotta, R. Hövel, M. Moser, S. Erhard, M. Karszewski, A. Giesen, and U. Keller, “16.2-W average power from a diode-pumped femtosecond Yb:YAG thin disk laser,” Opt. Lett. 25(11), 859–861 (2000). [CrossRef] [PubMed]
4. K. Beil, S. T. Fredrich-Thornton, C. Kränkel, K. Petermann, D. Parisi, M. Tonelli, and G. Huber, “New thin disk laser materials: Yb:ScYLO and Yb:YLF,” in CLEO/Europe and EQEC 2011 Conference Digest, OSA Technical Digest (CD) (Optical Society of America, 2011), paper CA11_6.
5. A. Pirri, D. Alderighi, G. Toci, M. Vannini, M. Nikl, and H. Sato, “Direct comparison of Yb3+:CaF2 and heavily doped Yb3+:YLF as laser media at room temperature,” Opt. Express 17(20), 18312–18319 (2009). [CrossRef] [PubMed]
6. 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(WO4)2 laser with single-pass pumping,” Opt. Lett. 33(7), 735–737 (2008). [CrossRef] [PubMed]
7. R. Gaumé, B. Viana, D. Vivien, J. P. Roger, and D. Fournier, “A simple model for the prediction of thermal conductivity in pure and doped insulating crystals,” Appl. Phys. Lett. 83(7), 1355–1357 (2003). [CrossRef]
8. M. Larionov, Kontaktierung und Charakterisierung von Kristallen für Scheibenlaser (Herbert Utz Verlag, 2009).
9. M. Larionov, K. Schuhmann, J. Speiser, C. Stolzenburg, and A. Giesen, “Nonlinear decay of the excited state in Yb:YAG,” in Advanced Solid-State Photonics, Technical Digest (Optical Society of America, 2005), paper TuB49.
10. K. Ueda, J. F. Bisson, H. Yagi, K. Takaichi, A. Shirakawa, T. Yanagitani, and A. Kaminskii, “Scalable ceramic lasers,” Laser Phys. 15(7), 927–938 (2005).
11. J. F. Bisson, D. Kouznetsov, K. Ueda, S. T. Fredrich-Thornton, K. Petermann, and G. Huber, “Switching of emissivity and photoconductivity in highly doped Yb3+:Y2O3 and Lu2O3 ceramics,” Appl. Phys. B 90, 201901 (2007).
12. C. Brandt, S. T. Fredrich-Thornton, K. Petermann, and G. Huber, “Photoconductivity in Yb-doped oxides at high excitation densities,” Appl. Phys. B 102(4), 765–768 (2011). [CrossRef]
13. A. Pirri, G. Toci, D. Alderighi, and M. Vannini, “Effects of the excitation density on the laser output of two differently doped Yb:YAG ceramics,” Opt. Express 18(16), 17262–17272 (2010). [CrossRef] [PubMed]
14. H. Nakao, A. Shirakawa, K. Ueda, H. Yagi, and T. Yanagitani, “CW and mode-locked operation of Yb3+-doped Lu3Al5O12 ceramic laser,” Opt. Express 20(14), 15385–15391 (2012). [CrossRef] [PubMed]
15. K. Beil, S. T. Fredrich-Thornton, F. Tellkamp, R. Peters, C. Kränkel, K. Petermann, and G. Huber, “Thermal and laser properties of Yb:LuAG for kW thin disk lasers,” Opt. Express 18(20), 20712–20722 (2010). [CrossRef] [PubMed]
16. A. Brenier, Y. Guyot, H. Canibano, G. Boulon, A. Rodenas, D. Jaque, A. Eganuan, and A. G. Petrosyan, “Growth, spectroscopic, and laser properties of Yb3+-doped Lu3Al5O12 garnet crystal,” J. Opt. Soc. Am. B 23(4), 676–683 (2006). [CrossRef]
17. J. Dong, K. Ueda, and A. A. Kaminskii, “Laser-diode pumped efficient Yb:LuAG microchip lasers oscillating at 1030 and 1047 nm,” Laser Phys. Lett. 7(10), 726–733 (2010). [CrossRef]
18. J. He, X. Liang, J. Li, H. Yu, X. Xu, Z. Zhao, J. Xu, and Z. Xu, “LD pumped Yb:LuAG mode-locked laser with 7.63ps duration,” Opt. Express 17(14), 11537–11542 (2009). [CrossRef] [PubMed]
19. A. Pirri, M. Vannini, V. Babin, M. Nikl, and G. Toci, “CW and quasi-CW laser performance of 10at.% Yb3+:LuAG ceramic,” Laser Phys. 23(9), 095002 (2013). [CrossRef]
20. M. Nikl, A. Yoshikawa, and T. Fukuda, “Charge transfer luminescence in Yb3+-containing compounds,” Opt. Mater. 26(4), 545–549 (2004). [CrossRef]
21. M. Kucera, M. Nikl, M. Hanus, Z. Onderisinova, and A. Beitlerova, “Growth, emission and scintillation properties of Tb-Sc doped LuAG epitaxial films,” IEEE Trans. Nucl. Sci. 59(5), 2275–2280 (2012). [CrossRef]
22. S. Chénais, F. Balembois, F. Druon, G. Lucas-Leclin, and P. Georges, “Thermal lensing in diode-pumped Ytterbium lasers—Part I: theoretical analysis and wavefront measurements,” IEEE J. Quantum Electron. 40(9), 1217–1234 (2004). [CrossRef]
23. L. Esposito, T. Epicier, M. Serantoni, A. Piancastelli, D. Alderighi, A. Pirri, G. Toci, M. Vannini, S. Anghel, and G. Boulon, “Integrated analysis of non-linear loss mechanisms in Yb:YAG ceramics for laser applications,” J. Eur. Ceram. Soc. 32(10), 2273–2281 (2012). [CrossRef]
