Tunable Laser Cavity. EM: End Mirror (flat); FM: Folding Mirror (radius of curvature 155 mm); OC: Output Mirror (flat); C denotes the direction of the crystal optical axis; P: tuning prism; S: slit. The inset shows the non-tunable cavity configuration used to measured the fraction of the pump power absorbed from crystals. M: flat mirror, F: pump beam filter, L: convergent lens 

Contexts in source publication

Context 1

... without a significant reduction of the emission lifetime. The main drawback in Yb-doped materials is a not negligible absorption at the lasing wavelength due to the thermal population of the lower levels of the laser transition. This requires a high pump intensity to reach the laser threshold. Several host material (e.g. oxides, fluorides, vanadates, borates or tungstates) are reported in literature, each of them with some particular and complementary characteristics. The direct comparison between the literature data regarding the laser performances achievable with different hosts materials and different doping levels is somehow complicated by the differences in the various experimental set-up, such as variations in the pump wavelength and geometry, cavity schemes, crystal cooling systems, mirrors reflectivity spectra, tuning devices. The aim of this study is to provide a direct comparison of the laser performances with the same experimental set-up to reduce the influences of the apparatus differences of two 3+ 3+ fluorinated crystals, namely Yb :CaF 2 [1,3,4] and Yb :YLF [ 2,5–7], to establish a clear comparison of the achievable laser performances. These materials are very interesting because of their very wide tuning range as well as for their long upper level life-time (>2 ms), which is much higher respect to the majority of the mostly used Yb doped hosts (see for instance [8] and references therein). This makes them very attractive for amplifier laser systems and generation of ultrashort pulses [11]. Therefore, the experiments were carried out on samples having similar geometries, with the same cavity layout and elements, and pumping device. Great attention was placed in the assessment of the effect of the pump-induced thermal load on the laser performance, by analyzing the behavior of the energy conversion and slope efficiency of the two laser devices under CW and quasi-CW pumping mode, with various Duty Factors (DF). Finally, we present the explored tunability range. Among the fluoride host crystals, CaF 2 shows good thermal properties [9]. The incomplete 3+ 2+ charge compensation of the dopant (Yb ) and the smaller replaced ion diameter (Ca ), hampers the use of high doping levels. The rearrangement of the crystal structure for the compensation of the charge defect results in a diversified site structure, which leads to a broader band emission fluorescence. This in turns determines a wide tunability [9], but on the other hand it determines the formation of clusters which can produce alternative de-excitation paths. Moreover, CaF is easily fabricated in bulk by standard method, i.e. Bridgman and Czochralski, or in thin films by Molecular Beam Epitaxy [10]. Due to its optical isotropy a simpler laser cavity is required. Conversely YLF allows for heavy Yb doping because of the exact ion charge substitution. The upper state lifetime is comparable with CaF 2 , i.e. 2 ms [8]. Its thermal conductivity is 3+ lower than that of the CaF 2 , and it further decreases for increasing Yb doping, see [12]. Concerning the polarization properties, YLF is an uniaxial crystal which permits two possible polarization states, i.e. π -pol and σ -pol, with different absorption and emission cross-section spectra. A preliminary investigation of the spectroscopic and laser properties of heavily doped samples, performed by our group, is reported elsewhere [6]. As for the pumping wavelength and geometry, both crystals feature a moderate and relatively constant value of absorption cross section around in the 920-950 nm window. With heavily doped Yb:YLF it is possible to obtain high absorption constants (of the order of 10-20 − 1 cm ), well suited for longitudinal pumping of the short (few mm) crystals with fiber-coupled diode lasers; conversely, with CaF 2 (which allows only for moderate doping) it is more difficult to obtain a strong pump absorption over a short crystal length in this wavelength region. Both Yb:YLF and Yb:CaF 2 can be pumped on the zero-phonon line, respectively at 958 and 976 nm. The Table 1 resumes the main spectroscopic and thermal properties of these crystals. Crystal growth of Yb-doped CaF 2 was performed in a vacuum-tight Czochralski system equipped with an automatic diameter control system. The resistive heater and thermal insulators were made of high-purity graphite. The starting materials were prepared from high- purity commercial (Stella Chimifa, Japan) fluoride powders of CaF 2 and YbF 3 (>99.99%). The concentration YbF 3 in the starting material was 5 mol%. For further details about the CaF 2 crystal growth, see [15]. 3+ The Yb :YLF lattice investigated in the experiment has been well described in [6]. Briefly, it was grown by a Czochralski technique employing LiF and YF 3 powders (from AC Materials, Tampa, Fla, Usa) as raw material for the lattice and adding a proper amount of YbF (>99.99%) powder to achieve a concentration of ytterbium as high as 30% at. The experimental apparatus is sketched in Fig. 1. The laser cavity is V-shaped with a folding half angle of 10° and with arm lengths of 78 mm (between the mirrors FM and EM) and 430 mm between the mirrors FM and OC. We have experimentally tested several values of this length ( i.e. between 200 and 600 mm) and we have found that for both crystals the laser output has only a weak dependence from the arm length. The chosen value maximizes the output power for both crystals. Several flat output couplers, OC, with transmission ranging from 1.5% to 20% were used. The crystal was carefully oriented with the facets perpendicular to the cavity axis, in order to re-inject the Fresnel reflection of the uncoated crystal faces into the cavity itself. The crystals (1.5 mm length for YLF, 2 mm for CaF 2 ) are welded with Indium on a copper heat sink. The heat sink is cooled with a Peltier device and stabilized at 18°C. The pump source is a laser diode emitting at 940 nm, coupled into an unpolarised fiber with 200 μ m core diameter and a numerical aperture of 0.44 (full angle). The measured pump intensity distribution in the focal plane results about Gaussian with a spot radius around 150 μ m @ 1/e . The laser was tuned by placing a tuning prism made of SF10 glass, with an apex angle of 60 degrees into a cavity arm between the output coupler and the folding mirror. The laser emission is tuned by tilting the OC-mirror around an axis perpendicular to the prism dispersion plane. The emission wavelength was measured with a fiber coupled, 60 cm focal length spectrometer equipped with a multichannel detector (spectral resolution 0.4 nm). In quasi-three level systems the pump absorption under high intensity levels and in lasing condition can be different from the linear ( i.e. low intensity) absorption, due to the combination of two competing effects: the saturation of the absorption because of the depletion of the lower laser level and the population draining from the upper laser level caused by the laser action [10]. Moreover the heating of the pumped volume can modify the thermal population distribution in the lower manifold. On the other hand the accurate assessment of the crystal absorption is needed for the determination of the laser threshold and efficiency. To clarify these aspects, we have investigated the behavior of the crystals absorption at high pump intensities, when the laser action is on and off, and under different thermal load conditions. In absence of the laser action, we have measured the fraction of power absorbed by the crystal for different DF (from 20% to 100%), with the same pump power range and focusing conditions used in the laser experiments. The measurement of the absorbed pump power when laser action is switched on was obtained by measuring the residual pump power transmitted through the crystal. To carry out this measurement, we modified the non tunable cavity of Fig. 1 by adding a flat folding mirror in the arm between the FM and the OC mirrors, which allowed the extraction of the pump fraction transmitted by the crystal and reflected by FM. The switching off of the laser action is achieved by obstructing the cavity arm between the folding mirror and the output coupler. The residual pump beam is collected by a converging lens and focused on a power meter. A short pass filter rejects the small amount of the laser radiation that leaks from the coating of the FM. Figure 2(a) and 2(b) report the results obtained by using various output couplers. In non- lasing condition we observed a decrease of several percent of the absorbed pump peak power fraction for increasing incident peak power, due to the absorption saturation. In the case of CaF 2 the absorption decreases of about 10%, and for YLF the absorption decreases of about 7%. For both crystals, when the laser action is turned on the absorption increases to a level between the saturated absorption for the same incident pump peak power and the unsaturated absorption. 3+ For the Yb :YLF, the absorption in lasing conditions is closer to the unsaturated absorption, while it remains fairly constant when the pump peak power increases; for CaF 2 the absorption monotonically decreases for increasing pump level. This behavior can be clarified by considering the dependence of the saturated absorption in presence of both an intense pump and laser field. It can be shown [10] that if the laser intensity I L in the crystal is much higher than the saturation intensity at the laser wavelength λ , defined ...

Context 2

... is comparable with CaF 2 , i.e. 2 ms [8]. Its thermal conductivity is 3+ lower than that of the CaF 2 , and it further decreases for increasing Yb doping, see [12]. Concerning the polarization properties, YLF is an uniaxial crystal which permits two possible polarization states, i.e. π -pol and σ -pol, with different absorption and emission cross-section spectra. A preliminary investigation of the spectroscopic and laser properties of heavily doped samples, performed by our group, is reported elsewhere [6]. As for the pumping wavelength and geometry, both crystals feature a moderate and relatively constant value of absorption cross section around in the 920-950 nm window. With heavily doped Yb:YLF it is possible to obtain high absorption constants (of the order of 10-20 − 1 cm ), well suited for longitudinal pumping of the short (few mm) crystals with fiber-coupled diode lasers; conversely, with CaF 2 (which allows only for moderate doping) it is more difficult to obtain a strong pump absorption over a short crystal length in this wavelength region. Both Yb:YLF and Yb:CaF 2 can be pumped on the zero-phonon line, respectively at 958 and 976 nm. The Table 1 resumes the main spectroscopic and thermal properties of these crystals. Crystal growth of Yb-doped CaF 2 was performed in a vacuum-tight Czochralski system equipped with an automatic diameter control system. The resistive heater and thermal insulators were made of high-purity graphite. The starting materials were prepared from high- purity commercial (Stella Chimifa, Japan) fluoride powders of CaF 2 and YbF 3 (>99.99%). The concentration YbF 3 in the starting material was 5 mol%. For further details about the CaF 2 crystal growth, see [15]. 3+ The Yb :YLF lattice investigated in the experiment has been well described in [6]. Briefly, it was grown by a Czochralski technique employing LiF and YF 3 powders (from AC Materials, Tampa, Fla, Usa) as raw material for the lattice and adding a proper amount of YbF (>99.99%) powder to achieve a concentration of ytterbium as high as 30% at. The experimental apparatus is sketched in Fig. 1. The laser cavity is V-shaped with a folding half angle of 10° and with arm lengths of 78 mm (between the mirrors FM and EM) and 430 mm between the mirrors FM and OC. We have experimentally tested several values of this length ( i.e. between 200 and 600 mm) and we have found that for both crystals the laser output has only a weak dependence from the arm length. The chosen value maximizes the output power for both crystals. Several flat output couplers, OC, with transmission ranging from 1.5% to 20% were used. The crystal was carefully oriented with the facets perpendicular to the cavity axis, in order to re-inject the Fresnel reflection of the uncoated crystal faces into the cavity itself. The crystals (1.5 mm length for YLF, 2 mm for CaF 2 ) are welded with Indium on a copper heat sink. The heat sink is cooled with a Peltier device and stabilized at 18°C. The pump source is a laser diode emitting at 940 nm, coupled into an unpolarised fiber with 200 μ m core diameter and a numerical aperture of 0.44 (full angle). The measured pump intensity distribution in the focal plane results about Gaussian with a spot radius around 150 μ m @ 1/e . The laser was tuned by placing a tuning prism made of SF10 glass, with an apex angle of 60 degrees into a cavity arm between the output coupler and the folding mirror. The laser emission is tuned by tilting the OC-mirror around an axis perpendicular to the prism dispersion plane. The emission wavelength was measured with a fiber coupled, 60 cm focal length spectrometer equipped with a multichannel detector (spectral resolution 0.4 nm). In quasi-three level systems the pump absorption under high intensity levels and in lasing condition can be different from the linear ( i.e. low intensity) absorption, due to the combination of two competing effects: the saturation of the absorption because of the depletion of the lower laser level and the population draining from the upper laser level caused by the laser action [10]. Moreover the heating of the pumped volume can modify the thermal population distribution in the lower manifold. On the other hand the accurate assessment of the crystal absorption is needed for the determination of the laser threshold and efficiency. To clarify these aspects, we have investigated the behavior of the crystals absorption at high pump intensities, when the laser action is on and off, and under different thermal load conditions. In absence of the laser action, we have measured the fraction of power absorbed by the crystal for different DF (from 20% to 100%), with the same pump power range and focusing conditions used in the laser experiments. The measurement of the absorbed pump power when laser action is switched on was obtained by measuring the residual pump power transmitted through the crystal. To carry out this measurement, we modified the non tunable cavity of Fig. 1 by adding a flat folding mirror in the arm between the FM and the OC mirrors, which allowed the extraction of the pump fraction transmitted by the crystal and reflected by FM. The switching off of the laser action is achieved by obstructing the cavity arm between the folding mirror and the output coupler. The residual pump beam is collected by a converging lens and focused on a power meter. A short pass filter rejects the small amount of the laser radiation that leaks from the coating of the FM. Figure 2(a) and 2(b) report the results obtained by using various output couplers. In non- lasing condition we observed a decrease of several percent of the absorbed pump peak power fraction for increasing incident peak power, due to the absorption saturation. In the case of CaF 2 the absorption decreases of about 10%, and for YLF the absorption decreases of about 7%. For both crystals, when the laser action is turned on the absorption increases to a level between the saturated absorption for the same incident pump peak power and the unsaturated absorption. 3+ For the Yb :YLF, the absorption in lasing conditions is closer to the unsaturated absorption, while it remains fairly constant when the pump peak power increases; for CaF 2 the absorption monotonically decreases for increasing pump level. This behavior can be clarified by considering the dependence of the saturated absorption in presence of both an intense pump and laser field. It can be shown [10] that if the laser intensity I L in the crystal is much higher than the saturation intensity at the laser wavelength λ , defined ...