Open Access
Add to BibSonomy
Abstract
The influence of rare earth dopant concentration on selenide laser glass quality was investigated. A problem to be solved was identified – an increase in rare earth doping level leads to optical losses due to light scattering by heterogeneous inclusions and to the decrease of the optical damage threshold. The room temperature laser system consisting of 4.6 μm pulsed Fe:ZnSe pump laser and Ce3+:Ge20Sb10Ga5Se65 bulk glass was demonstrated. The slope efficiency of Ce3+:glass laser with respect to the absorbed pump energy reached 25% with a maximum output energy of 45 mJ.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Recent achievements in the synthesis of high-purity chalcogenide glasses and doping them by the rare earth elements (REE) have made a great impact on the solid-state lasers that can operate in the spectral region from 5 to 6 $\mu$m [1–3]. A laser action was obtained in Tb$^{3+}$ - doped [4,5], Pr$^{3+}$- doped [6] and Ce$^{3+}$ - doped [7,8] selenide glasses and fibers. In [5] a selenide glass fiber (core - Ge$_{20}$Ga$_5$S$_{10}$Se$_{65}$ doped with 2$\times$10$^{19}$ cm$^{-3}$ of Tb$^{3+}$ ions; cladding –Ge$_{12}$As$_{20}$Sb$_5$S$_{63}$) was used as the active fiber for the laser. A continuous laser action at room temperature (RT) under 2 $\mu$m pumping at spectral band of 5.20 – 5.28 $\mu$m and a slope efficiency of 5.3% was demonstrated. In [9] a step-index selenide-chalcogenide fiber was used as an active medium with core composition of 500 ppmw Ce - Ge$_{15}$As$_{21}$Ga$_1$Se$_{63}$ and cladding composition of Ge$_{21}$Sb$_{10}$Se$_{69}$. The continuous wave (CW) room temperature pump source was a 4.15 $\mu$m quantum cascade laser. The authors are claiming for the CW laser operation in the wavelength range from 5.13 $\mu$m to 5.28 $\mu$m, but unfortunately the total output power of the Ce-doped fiber laser was less than 100 $\mu$W and there is no data on the laser efficiency. In contrast, major laser characteristics have been demonstrated in [7] with a bulk glass with a composition of Ge$_{20}$Sb$_{10}$Ga$_5$Se$_{65}$ doped with 3$\times$10$^{19}$ cm$^{-3}$ of Ce$^{3+}$ ions. The pump source was a liquid nitrogen cooled pulsed Fe:ZnSe laser operating at 4.1 $\mu$m. Although that work marked a breakthrough in Ce-doped glass lasers, but the efficiencies of these lasers remain rather far from the Stokes shift theoretical limit; currently, the maximum obtained slope efficiency with respect to the absorbed pump energy was about 21% [8]. At the same time, the slope efficiency with respect to the total energy of the pumping laser was 10% only, since the pump wavelength did not coincide with the maximum of the Ce$^{3+}$ absorption band and about 20% of the pumping energy was lost due to Fresnel losses on the input face of the active element.
Another parameter to be optimized is the loss of the laser radiation in the active glass caused by the inclusions and by a parasitic absorption by Se-H and Ge-H groups having characteristic absorption bands at 4.4 and 4.9 $\mu$m, respectively. The problem is further complicated by the circumstance that both the scattering and absorption losses can be strongly dependent on the rare earth doping level in the laser glass. This issue is of big practical importance but remains poorly investigated. In particular, cross-relaxation [10] and sensitization [11] pumping schemes inevitably require rather high ($\sim$10$^{20}$ cm$^{-3}$) concentrations of rare earth dopants. Therefore, there is still a great potential for efficiency improvements of Ce-doped glass lasers, which we address in the current work.
In this paper, we demonstrate multiple improvements in comparison with [7,8]. Firstly, we experimentally investigate the relationship between the magnitude of optical losses and the doping level of the glass. Two samples of a selenide glass were under consideration: a sample 1 simultaneously doped by two active ions, Dy$^{3+}$ and Ce$^{3+}$, and a sample 2 doped only with Ce$^{3+}$. Secondly, we focus on the improvement of the parameters (efficiency and RT operation) of the pulsed Ce$^{3+}$ - doped selenide bulk glass laser at 5.2 $\mu$m. Two modifications of the pumping scheme were considered: 1. the usage of the process of excitation transfer from Dy$^{3+}$ to Ce$^{3+}$ ions, when a RT pulsed Er:YAG laser emitting at 1.78 $\mu$m matching with the dysprosium absorption band was used as a pump source of the sample 1; 2. the usage of a RT 4.6 $\mu$m Fe:ZnSe laser for the cerium in-band pumping of both samples. A thermally induced lens dynamics in the active selenide glass is also discussed.
2. Preparation and properties of the Ce$^{3+}$-doped glass samples
The glass samples used in our experiments (base composition Ge$_{20}$Sb$_{10}$Ga$_5$Se$_{65}$) have been synthesized in identical conditions using the technology described in [1,12]. At the first stage, a high-purity charge containing Ge, Sb, and Se in a given ratio was synthesized. Then, CeI$_3$ was prepared under high vacuum conditions and loaded into the charge together with a given amount of gallium by the chemical transport reactions method [12]. The ampule with the resulting charge was melted in a rocking furnace for 5 h at 850 $^\circ$C, quenched in air, and annealed at 310 $^\circ$C. We prepared two active glass samples. The sample 1 was co-doped with Dy$^{3+}$ (nominal concentration of 1$\times$10$^{20}$ cm$^{-3}$) and Ce$^{3+}$ (nominal concentration 3$\times$10$^{19}$ cm$^{-3}$) while the sample 2 was doped with Ce$^{3+}$ only at the nominal concentration of 3.1$\times$10$^{19}$ cm$^{-3}$.
The macrocomposition of the glass and the REE content were determined by the ICP-AES method [13]. Obtained values Ge$_{20.4\pm 0.1}$Sb$_{10.3 \pm 0.1}$Ga$_{4.8 \pm 0.1}$Se$_{64.5 \pm 0.1}$, (3.1$\pm$0.1)$\times$10$^{19}$ cm$^{-3}$ Ce, and (1$\pm$0.1)$\times$10$^{20}$ cm$^{-3}$ Dy correspond well to the nominal. The content of metal impurities Al, Ba, Ca, Co, Cr, Cu, Fe, K, Mg, Mo, Na, Ni, Pb, Sn, Ta, Ti, V, W, Zn in the glass was below the detection limit of the ICP-MS method (0.002–0.5 ppm(wt)), Mn – 0.03 ppm(wt) (ELEMENT 2 high resolution mass spectrometer, ThermoFinnigan). The content of heterogeneous inclusions larger than 1 $\mu$m in glass, according to optical microscopy, was below the detection limit of 10$^2$ cm$^{-3}$ (Axio Imager.M2m microscope). The concentration of hydrogen impurity in the form of Se-H and Ge-H groups, which has an absorption bands near 5 $\mu$m, in chalcogenide glasses prepared using these methods is at the level of 0.01 ppm(wt) (1 ppm(at)) and lower [1].
The attenuation spectra of two synthesized glasses (Fig. 1) were measured using a set of cylindrical rods of different lengths. The attenuation spectrum of the sample 1 demonstrates intensive Dy$^{3+}$ absorption bands at 1.7 $\mu$m and 3 $\mu$m. The wide 3.6-6 $\mu$m absorption band in the spectra of both samples corresponds to Ce$^{3+}$ ions. It was impossible to detect the weak absorption bands of residual Se-H and Ge-H groups in the spectral range of 4-5 microns against the background of this intense band. The low enough concentration of these impurities in the investigated samples is verified by the high Ce$^{3+}$ luminescence lifetimes (3.7 ms in both samples). It can be seen, that there is a difference in the absorption coefficient between glass samples in the range 1.5–3.5 $\mu$m and almost equal within spectral range of 6–8 $\mu$m. During high-temperature homogenizing melting of the charge, rare earth iodides can interact with the silica glass ampule material [14]. This leads to contamination of chalcogenide glass with SiO$_2$ admixture in the form of submicron inclusions. Absorption bands of Si-O bonds at about 9.3 $\mu$m are clearly observed in the IR absorption spectra of both samples (Fig. 1). Luckily, this wavelength lies far from absorption and emission of Ce$^{3+}$ ions. The intensity of the 9.3 $\mu$m absorption peak is proportional to the volume fraction of SiO$_2$, but not necessarily to the number of heterogeneous inclusions. Nevertheless, the presence of even submicron inclusions of any nature can reduce the optical damage threshold, cause spectrally non-selective scattering losses and thus affect the lasing properties of the glass [15]. The evaluation of these losses is discussed below.
Two cylindrical glass rods were prepared with a diameter of 12 mm. The length of the sample 1 was 24 mm, and 22 mm of the sample 2. Both end-surfaces of the rods were polished to an optical quality with a wedge less than 15 arc seconds. The working surfaces of both samples had no antireflection coatings. A photo of the both active elements is shown in Fig. 2(a).
Fig. 2. a) A photo of the Dy$^{3+}$-Ce$^{3+}$-glass (sample 1) and Ce$^{3+}$-glass (sample 2) active elements; b) transmission spectra of the samples.
The transmission spectra (Fig. 2(b)) of both samples were recorded using Fourier-transform spectrometer. The absorption of both samples near the maximum of the Ce$^{3+}$ ion absorption band was about 90% (with respect to the radiation entering the sample).
Before the lasing experiments a comparative study on the light scattering has been carried out by measuring the level of the scattered pump signal for both samples. In these experiments, a 4.6 $\mu$m pump beam with a diameter of 2 mm from Fe:ZnSe laser (described in section 4) was passed through the samples. The distance from the lateral cylindrical surface of the sample to the beam axis was 2 mm. The scattered radiation passed through the lateral surface and subsequently was recorded perpendicular to the direction of the pump beam propagation using InAsSb Amplified Photodetector PDA07P2 (Thorlabs Inc.). Measurements were performed at 110 mJ of incident Fe:ZnSe laser radiation. The obtained results are presented in Fig. 3. The upper traces (red) are the oscillograms of the Fe:ZnSe laser pulses. The lower traces (blue) are the signals registered by the InAsSb photodetector.
Fig. 3. Measured light scattering patterns for the glass samples (a – sample 1, b – sample 2). Upper traces (red) – pump pulse, lower traces (blue)– measured scattered signals.
For both samples, the measured shapes of the scattered signals demonstrate a gradual growth in the overall signal level due to appearance of the Ce$^{3+}$ luminescence around 5 $\mu$m and the spiky structure of the scattered pump pulse can be clearly seen. The incomplete correlation of the spikes in the scattered light signals to the spikes in the pump pulses is apparently explained by the multimode generation of Fe:ZnSe laser, in which the emergence of a new spike is accompanied by a "mode jump", i.e. a change in the spatial structure of the laser beam. In fact, different spikes pass through different sections of the sample, in which, due to the inhomogeneous distribution of scattering centers in the glass, the scattering can be very different. However, it is clearly seen that sample 2 shows 3-4 times lower scattering level compared to the sample 1. These qualitative data match with Fig. 1, where some optical losses (growing with the wavelength decrease) are clearly seen for the sample 1 at about 2 $\mu$m, and much less loss values are noticeable for the sample 2. It is also worth noting, that there is practically no correlation between the scattering losses and the intensity of Si-O absorption peaks.
3. Pulsed 1.78 $\mu$m pumping of the Dy$^{3+}$-Ce$^{3+}$-doped glass sample
In this section, we used a RT 250-$\mu$s pulsed Er:YAG (Er$^{3+}$ concentration of 5%) laser at 1.78 $\mu$m as a pump source to achieve laser action of Ce$^{3+}$ in glass sample 1 due to the transfer of excitation from Dy$^{3+}$ ions to Ce$^{3+}$ ions. This wavelength was chosen according to the fact that Ce$^{3+}$ luminescence rise time is the shortest when pumping into $^64$H$_{11/2}$ Dy$^{3+}$ state (Fig. 4(a)) due to good resonance of the interacting transitions [11]. The experimental setup is shown in Fig. 4(b). The pump beam was launched to the active element at an angle of 3$^\circ$ to the 95-mm-long cavity that was formed by curved aluminum coated mirror (R100) and output coupling (OC) mirror with 2% transmission in the vicinity of 5.2 $\mu$m.
Fig. 4. a) Energy levels of the Dy and Ce ions and possible energy transfer. b) The experimental setup with 1.78 $\mu$m pump source.
Unfortunately, we did not observe any lasing in this configuration under pump energy as high as 250 mJ and pump beam spot of 1.5 mm on the active glass. There can be several reasons for it: 1. The Dy$^{3+}\to$ Ce$^{3+}$ energy transfer process shown in Fig. 4(a) may be reversible because the lifetime of Dy$^{3+}$ $^64$H$_{13/2}$ level at the chosen active ions concentrations remains high (1300 $\mu$s [11]) and the transitions involved in the process are in good resonance. The mechanisms of such reverse energy transfer may be either radiationless or radiative. The latter means the possibility of excited state absorption at the lasing wavelength. Still the sensitization scheme proposed in [11] has not exhausted its capabilities, and in case of accurate optimization of active ions concentrations it may work out. 2. The additional optical losses in the glass with the enhanced overall concentrations of rare earth ions can cause the increased amount of light scattering, as was shown in the previous section. The evaluation of the passive losses for two active glass samples at the lasing operation will be presented in the following section.
4. Pulsed laser operation of the glass samples under 4.6 $\mu$m pumping
The experimental setup of the room temperature laser system is shown in Fig. 5. A homemade RT 200-$\mu$s pulsed Fe:ZnSe laser at a wavelength of 4.6 $\mu$m excited by an 2.94 $\mu$m Er:YAG (Er$^{3+}$ concentration of 50%) laser was used as a pump source for the Ce$^{3+}$-doped chalcogenide glass. In contrast to [7,8], where the Fe:ZnSe crystal was installed inside a vacuum chamber and cooled with liquid nitrogen, in this work the crystal was located in the atmosphere and cooled by the Peltier element to $\sim$9$^\circ$C. The Fe:ZnSe laser cavity was formed by a plane OC mirror with 25% transmission at 4.6 $\mu$m and curved gold coated mirror with 500 mm radius of curvature. The cavity length was 170 mm. The Fe:ZnSe crystal was placed 25 mm from the OC mirror. Maximum pulse energy that was obtained from the Fe:ZnSe laser was about 350 mJ. The laser could operate at a repetition rate of 0.2 Hz for 2 minutes, which was limited by the heating capacity of the Fe:ZnSe crystal cooling system. The Fe:ZnSe optical spectrum measured at 200 mJ of output energy is shown in Fig. 6(a). It can be seen that the Fe:ZnSe laser spectrum lies in the vicinity of the maximum of the absorption band of Ce$^{3+}$ in the glass. The pump radiation from RT Fe:ZnSe laser was directed with a 750 mm radius of curvature aluminum coated mirror onto the Ce$^{3+}$-doped glass active element at an angle of about 5$^\circ$ to the laser cavity axis. The diameter of the pump beam inside the active element could vary depending on the distance from the focusing mirror to the active element. For more efficient use of pumping, the radiation of Fe:ZnSe laser reflected from the input surface of the active element was returned to the active volume using an aluminum spherical (R=100 mm) mirror that was placed at the distance about 100 mm from the active element.
Fig. 6. a) Fe:ZnSe laser output spectrum; b) Oscilloscope traces of the Er:YAG, Fe:ZnSe and Ce$^{3+}$:glass lasers; c) Ce$^{3+}$:glass laser output spectrum for 43% transmission of the OC.
The Ce$^{3+}$-glass lasing was investigated for two diameters of the pump spot of 2.5 and 1.2 mm. A 140-mm-long Ce$^{3+}$-doped glass laser cavity was composed by a 300-mm radius of curvature spherical aluminum mirror and a plane dielectric OC mirror. The uncooled cylindrical laser element was glued to a flat 7$\times$7 mm aluminum base at a distance of 8 mm from the OC mirror. The input-output characteristics of the laser were studied using OC with transmission 43% which was found to be optimal in previous studies [8]. For Caird-plot analysis three OCs with transmission of 1.5%, 4.5%, 23% were additionally used. The typical oscilloscope traces of the Er:YAG, Fe:ZnSe and Ce$^{3+}$-doped glass lasers are demonstrated in Fig. 6(b). The output optical spectrum of the cerium laser measured at 6 mJ of output energy and 43% OC is shown in Fig. 6(c). The structure in the spectrum is due to interference effects on the resonator elements and absorption lines of atmospheric air [8].
When pumped into the Ce$^{3+}$ absorption band, generation on Ce$^{3+}$ ions was easily obtained in both samples. The efficiency curves in terms of the output energy with respect to the absorbed pump energy for two pump spot diameters are shown in Fig. 7. The measurements were carried out in the single-pulse mode of laser operation.
In the case of 2.5 mm pump spot, the input-output curves demonstrate a liner growth within the entire range of pumping energies and the optical-to-optical slope efficiency with respect to the absorbed pump energy was measured to be 25% and 20% for the sample 2 and sample 1, respectively, which we attribute to the higher optical quality of sample 2. The slope efficiency with respect to the total pumping energy of 18% was achieved for the sample 2. The maximum output energy was of 45 mJ. For smaller pump beam diameter the laser demonstrates lower threshold, but slightly reduced slope efficiency what we associate with the less overlap of the pump beam and the laser mode. As can be seen from Fig. 7, the quenching of the lasing radiation is appearing starting from about 40 mJ of the absorbed pump energy for smaller pump beam diameter. This effect can be associated with thermal lensing which can be confirmed by the Ce$^{3+}$ laser output pulse shape dynamics. In the case of the 2.5 mm pump beam spot, no efficiency degradation in the range of absorbed pump energies from 100 to 270 mJ is observed. Figure 8 demonstrates several oscilloscope traces recorded at various pump conditions. It can be seen that at high pumping energy in a small spot (Fig. 8(b)), lasing terminates earlier than the pump pulse, unlike in the other cases (Figs. 8(a,c,d)). This, in our opinion, is due to the appearance of a dynamic thermal lens in the active volume, which makes the laser resonator unstable. As a result, the increase in the output energy of the laser is saturated.
Fig. 8. Temporal behavior of Fe:ZnSe laser pulses (upper) and Ce$^{3+}$:glass laser pulses (lower) for two pump beam spots at different pump energies entered the active element.
To estimate the potential limit of the thermal lens influence, we calculated the focal length of the thermal lens using the equation from [16]:
where $C$ is the heat capacity per unit volume, $\frac {dn}{dT}$ is the thermo-optic coefficient, $r_0$ – radius of the pump beam and $E$ is the energy that is spent on heating. We assume that for the Ge$_{20}$Sb$_{10}$Ga$_5$Se$_{65}$ glass thermo-optic coefficient can be estimated at the level of 80$\cdot$10$^{-6}$ K$^{-1}$ [17] and C = 1.38 J$\cdot$cm$^{-3}\cdot$K$^{-1}$ [18]. The calculated focal length for two pump beam diameters as a function of absorbed pump energy is shown in Fig. 9(a). The thermal lens influence on the laser operation can be clearly seen on the stability of the laser cavity that is presented as a beam waist diameter on the active element (see Fig. 9(a), red curve). From these results it can be seen that the thermal lens of 13 cm or less can be crucial for the laser performance. In the case of 2.5 mm pump beam the influence of the thermal lens can be expected at the absorbed pump power higher than 300 mJ. According to the calculation, for the 1.2 mm pump beam, thermal lens should be critical at absorbed pump energies of more than 15 mJ, which is not consistent with the experiment. However, it should be noted that the calculated values presented in Fig. 9(a) correspond to the end of the pump pulse when the focal length of the induced thermal lens reaches its minimum. It was also assumed that the laser beam axis coincides with the pump beam axis. In experiments, the angle between them was 5$^\circ$, and the laser beam crossed only part of the thermal lens area. Therefore, with a smaller pump spot, lasing occurred even at absorbed energies over 150 mJ, but its failure was observed before the end of the pump pulse.Fig. 9. a) Calculated focal length of the thermal lens in the Ce$^{3+}$:glass as a function of the absorbed pump energy (blue and green) and the laser mode diameter at active element versus focal length of the induced thermal lens (red); b) The surface damage of the Dy$^{3+}$-Ce$^{3+}$:glass sample 1.
Figure 7 also shows that with a pump spot of 1.2 mm, the saturation of the curve for the sample 1 occurs faster compared to sample 2, which, in our opinion, is due to the occurrence of damage inside the sample 1 under high pump energy. This was manifested by the fact that after irradiating the sample 1 with energies of more than 100 mJ, it was not possible to reproduce the values of the laser output energies, which were obtained at the corresponding pump energies before such exposure. In addition, at pump energies of more than 150 mJ, damage also occurred on the surface of the sample 1 (Fig. 9(b)). In the case of sample 2, no surface damage was observed in the studied pump energy range, and internal damage occurred at an energy of more than 200 mJ in the 1.2 mm spot, which corresponds to an energy density of 18 J/cm$^2$.
5. Internal losses of the active glass samples from the laser performance
One of the methods for measuring the resonator passive losses at the lasing wavelength is a well-known Caird analysis [19]. In this study, input-output laser characteristics of the both samples have been measured for OCs with transmission of 1.5%, 4.5%, 23% and 43%. As an example, the obtained efficiency curves for the sample 2 are presented in Fig. 10(a). Similar to [8], Caird plots (Fig. 10(b)) were generated for both active glass samples using the effective cavity transmissions $T_{eff}$ of the laser resonator (the transmission of the Fabry-Perot interferometer formed by the OC and active element’s end facet) and the measured slope efficiencies. As a result, the round-trip passive losses of 9% and 5% were retrieved for the laser resonators with sample 1 and sample 2 respectively, which also demonstrates the higher optical quality of the sample 2 material.
Fig. 10. a) Output energy of the Ce$^{3+}$:glass laser (sample 2) versus the absorbed pump energy with different OCs. b) Caird-plot: dependence of the inverse slope efficiency on the transmission of the Fabry-Perot cavity formed by the OC and active element’s end facet.
6. Induced thermal lens dynamics
All the measurements above have been conducted at a single-pulse operation regime as had been done in earlier studies [7,8]. It is important to know the possibilities of the Ce$^{3+}$:glass active element to operate in a pulse-periodic mode. For this purpose, we studied an induced thermal lens dynamics by measuring a probe beam signal of a 1550 nm laser diode, which is passing through the glass sample (without cavity mirrors). Here we used sample 2 as it shows better optical quality. The probe beam was collimated to a diameter of 2 mm and launched thought the center of the active glass sample at the same area where the pump beam was directed. We used the pump beam diameter of 2.5 mm in that case. The passed probe beam was monitored by an InGaAs photodetector with a sensor diameter of 0.5 mm, that was placed at 300 mm from the Ce-glass sample. The appearance of a thermal lens led to a change in the probe beam divergence and subsequently yields to the change of the probe beam spot size in the plane of the photodetector sensor, as a result of which the signal amplitude changed. Figure 11 shows the probe beam signal (blue) and the Fe:ZnSe laser pump pulse (orange) at different time scales. The measurements were carried out at the level of 110 mJ of absorbed pump energy.
Fig. 11. Thermal lens dynamics measurements: a) Probe signal at 1 ms time scale. b) Probe signal at 100 s time scale.
On a scale of 1 ms one can observe an increase of the probe signal due to appearance of the thermal lens. As was estimated earlier the focal length of the induced thermal lens at 110 mJ is longer than 300 mm, hence the probe beam is slightly focused after the active element and we observe rising of the probe signal at the photodetector immediately after pump pulse (Fig. 11(a)). On longer time scale of 100 seconds (Fig. 11(b)), we can see a slow recovery of the thermal lens due to heat dissipation. It should be noted, that no additional cooling system was applied to the Ce$^{3+}$:glass element. There was only small contact area between the cylindrical Ce$^{3+}$:glass rod and a 7$\times$7 mm flat aluminum base. The temperature in the lab was around 20$^\circ$C. The recovery of the probe signal is fitted by the exponential curve resulting in a time constant of 4.4 s. This result shows that the potential repetition rate of the Ce-doped glass laser can be estimated at the level about 0.2 Hz. We tested the operation of the laser when it was pumped with the same repetition rate of 0.2 Hz under a pump energy of 200 mJ and a pump beam diameter of 2.5 mm. During the 100 s series, which was limited by the capabilities of the Fe:ZnSe laser, we did not observe any decrease in the output energy, which was close to the laser energy in the mode of single pulses.
Since the characteristic spreading time of the thermal lens for the cylindrical geometry of the sample is proportional to the square of its radius, when using an active sample with a smaller diameter, an increase in the pulse repetition frequency can be expected. With proper additional cooling of the active element, this value can be further increased.
7. Conclusion
The influence of rare earth dopants concentration on the selenide laser glass quality was investigated. It was found that the glass doping with up to 1.3$\times$10$^{20}$ cm$^{-3}$ of rare earth ions (cerium and dysprosium) does not cause any noticeable mid-infrared Ce$^{3+}$ luminescence quenching. However, a problem to be solved was also identified - an increase in rare earth doping level leads to optical losses due to light scattering by heterogeneous inclusions and to the decrease of the optical damage threshold. The room temperature laser system consisting of 4.6-$\mu$m pulsed Fe:ZnSe pump laser and Ce$^{3+}$:glass rod on the base of Ge$_{20}$Sb$_{10}$Ga$_5$Se$_{65}$ glass is demonstrated. The slope efficiency of Ce$^{3+}$:glass laser with respect to the pump energy of 25% (18% with respect to the total pumping energy) is achieved. The output energy as high as 45 mJ is obtained.
Funding
Ministry of Science and Higher Education of the Russian Federation (FFSR-2021-0001).
Acknowledgments
The authors are acknowledge to V.G. Polushkin from the Engineering center for new technologies (Institute for Nuclear Research, Russian Academy of Sciences) for supplying the Er:YAG laser rods. We acknowledge the support by the Open Access Publication Fund of the University of Duisburg-Essen.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. A. Velmuzhov, M. Sukhanov, V. Plotnichenko, A. Plekhovich, V. Shiryaev, and M. Churbanov, “Preparation of REE-doped Ge-based chalcogenide glasses with low hydrogen impurity content,” J. Non-Cryst. Solids 525, 119669 (2019). [CrossRef]
2. M. F. Churbanov, B. I. Denker, B. I. Galagan, V. V. Koltashev, V. G. Plotnichenko, M. V. Sukhanov, S. E. Sverchkov, and A. P. Velmuzhov, “Comparison of 4.5–6 μm luminescent and lasing properties of rare earth dopants in chalcogenide glasses,” J. Lumin. 245, 118756 (2022). [CrossRef]
3. V. Nazabal and J.-L. Adam, “Infrared luminescence of chalcogenide glasses doped with rare earth ions and their potential applications,” Opt. Mater.: X 15, 100168 (2022). [CrossRef]
4. V. Shiryaev, M. Sukhanov, A. Velmuzhov, E. Karaksina, T. Kotereva, G. Snopatin, B. Denker, B. Galagan, S. Sverchkov, V. Koltashev, and V. Plotnichenko, “Core-clad terbium doped chalcogenide glass fiber with laser action at 5.38 μm,” J. Non-Cryst. Solids 567, 120939 (2021). [CrossRef]
5. B. Denker, B. Galagan, V. Koltashev, V. Plotnichenko, G. Snopatin, M. Sukhanov, S. Sverchkov, and A. Velmuzhov, “Continuous Tb-doped fiber laser emitting at ∼5.25 μm,” Opt. Laser Technol. 154, 108355 (2022). [CrossRef]
6. M. F. Churbanov, B. I. Denker, B. I. Galagan, V. V. Koltashev, V. G. Plotnichenko, G. E. Snopatin, M. V. Sukhanov, S. E. Sverchkov, and A. P. Velmuzhov, “Laser potential of Pr3+ doped chalcogenide glass in 5-6 μm spectral range,” J. Non-Cryst. Solids 559, 120592 (2021). [CrossRef]
7. B. I. Denker, B. I. Galagan, S. E. Sverchkov, V. V. Koltashev, V. G. Plotnichenko, M. V. Sukhanov, A. P. Velmuzhov, M. P. Frolov, Y. V. Korostelin, V. I. Kozlovsky, S. O. Leonov, P. Fjodorow, and Y. K. Skasyrsky, “Resonantly pumped Ce3+ mid-infrared lasing in selenide glass,” Opt. Lett. 46(16), 4002–4004 (2021). [CrossRef]
8. P. Fjodorow, M. P. Frolov, S. O. Leonov, B. I. Denker, B. I. Galagan, S. E. Sverchkov, V. V. Koltashev, V. G. Plotnichenko, M. V. Sukhanov, and A. P. Velmuzhov, “Mid-infrared laser performance of Ce3+-doped selenide glass,” Opt. Express 29(17), 27674–27682 (2021). [CrossRef]
9. J. J. Nunes, Ł. Sojka, R. W. Crane, D. Furniss, Z. Q. Tang, D. Mabwa, B. Xiao, T. M. Benson, M. Farries, N. Kalfagiannis, E. Barney, S. Phang, A. B. Seddon, and S. Sujecki, “Room temperature mid-infrared fiber lasing beyond 5 µm in chalcogenide glass small-core step index fiber,” Opt. Lett. 46(15), 3504–3507 (2021). [CrossRef]
10. M. F. Churbanov, B. I. Denker, B. I. Galagan, V. V. Koltashev, V. G. Plotnichenko, S. E. Sverchkov, M. V. Sukhanov, and A. P. Velmuzhov, “Peculiarities of 1.6-7.5μm Pr3+ luminescence in Ge36Ga5Se59 glass,” Opt. Mater. Express 9(11), 4154–4164 (2019). [CrossRef]
11. M. F. Churbanov, B. I. Denker, B. I. Galagan, V. V. Koltashev, V. G. Plotnichenko, M. V. Sukhanov, S. E. Sverchkov, and A. P. Velmuzhov, “Cascade sensitization of mid-infrared Ce3+ luminescence by Dy3+ ions in selenide glass,” J. Lumin. 231, 117809 (2021). [CrossRef]
12. A. Velmuzhov, M. Sukhanov, N. Zernova, V. Shiryaev, T. Kotereva, L. Ketkova, I. Evdokimov, and A. Kurganova, “Preparation of Ge20Se80 glasses with low hydrogen and oxygen impurities content for middle ir fiber optics,” J. Non-Cryst. Solids 521, 119505 (2019). [CrossRef]
13. I. Evdokimov, D. Fadeeva, A. Kurganova, V. Shiryaev, V. Pimenov, and E. Karaksina, “Determination of matrix elements and praseodymium in Ga–Ge–As–Se glasses by inductively coupled plasma–atomic emission spectrometry,” J. Anal. Chem. 75(7), 869–877 (2020). [CrossRef]
14. W. Erk, “The reaction of rare-earth metal iodides with the silica wall in metal halide discharge lamps,” Philips J. Res. 42, 102–118 (1987).
15. L. Ketkova and M. Churbanov, “Heterophase inclusions as a source of non-selective optical losses in high-purity chalcogenide and tellurite glasses for fiber optics,” J. Non-Cryst. Solids 480, 18–22 (2018). [CrossRef]
16. M. Frolov, Y. V. Korostelin, V. Kozlovsky, Y. P. Podmar’kov, and Y. K. Skasyrsky, “High-energy thermoelectrically cooled Fe:ZnSe laser tunable over 3.75–4.82 μm,” Opt. Lett. 43(3), 623–626 (2018). [CrossRef]
17. L. Bodiou, Y. Dumeige, S. Normani, G. Louvet, P. Němec, V. Nazabal, and J. Charrier, “Design of a multimode interferometer-based mid-infrared multispecies gas sensor,” IEEE Sens. J. 20(22), 13426–13435 (2020). [CrossRef]
18. J. D. Musgraves, S. Danto, and K. Richardson, “Thermal properties of chalcogenide glasses,” in Chalcogenide Glasses (Elsevier, 2014), pp. 82–112.
19. J. A. Caird, S. A. Payne, P. R. Staver, A. Ramponi, and L. Chase, “Quantum electronic properties of the Na3Ga2Li3F12: Cr3+ laser,” IEEE J. Quantum Electron. 24(6), 1077–1099 (1988). [CrossRef]
