Open Access
Add to BibSonomy
Abstract
We investigate the cubic sesquioxide Dy3+:Lu2O3 as a gain material for lasers in the mid-infrared spectral range. Bulk crystals of this material have been grown from the melt for the first time in the course of our experiments. Spectroscopic characterizations of the material show that it compares very favorably with more established materials like Dy3+:ZBLAN. The energetic positions of the energy levels up to 14,000 cm−1 were identified by spectroscopy at 11 K. Additionally, absorption cross-sections as high as 2.2 × 10−20 cm2 at 1.26 µm, and emission cross-sections as high as 6.7 × 10−21 cm2 at 2.777 µm, were determined. Furthermore, the room temperature fluorescence lifetime of the 6H13/2 multiplet was determined as 50 µs, which points to serious quenching as compared to the estimated radiative lifetime in the order of 20 ms.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Devices relying on coherent mid-infrared light sources are found, among others, in medical, metrological, remote sensing, and THz generation implementations [1–3]. Mid-infrared laser systems have mostly relied on low-phonon-energy fluorides or chalcogenides [4]. In recent years, mid-infrared lasers realized with Dy3+ as the active ion have exclusively relied on ZBLAN glasses [5–7] as a host material. These can be delicate to handle and suffer from a very low thermal conductivity of just 0.6 Wm−1K−1 [8]. Single-crystalline oxide host materials on the other hand can provide good thermomechanical properties but suffer from high phonon energies [9], which can lead to an increased rate of non-radiative decay for low-energy mid-infrared transitions.
Cubic sesquioxide host materials, in which the ratio of oxygen ions to metal ions is three to two and which belong to the space group , feature very good thermooptical and -mechanical properties. Their thermal conductivity can reach up to 18 Wm−1K−1 [10]. At the same time, their maximum phonon energies are comparable to those of ZBLAN, in the order of 600 cm−1 [11,12]. This makes them interesting host materials for mid-infrared laser applications, even though their growth is challenging due to their high melting points of about 2400 °C [12]. Over the last years impressive laser results have been obtained from the near- to the mid-infrared spectral range [13–15]. No mid-infrared laser action has been reported so far with any oxide host material utilizing Dy3+ as the active ion. As a host material Lu2O3 would be an ideal first candidate due to the similarity in ionic radii of Lu3+ and Dy3+ leading to a homogenous distribution of Dy3+ throughout the crystal. The crystal structure is simple cubic and the lattice constant is 10.384 Å [16]. In addition, the thermal conductivity of Lu2O3 is largely invariant regarding doping with heavy rare-earth ions [10]. Therefore, this material has also recently been grown by the optical floating zone method and characterized in the visible spectral range in regards to the possibility as a gain material for yellow lasers [17].
Here, in a continuation of our previous work [18] we report in more detail on the first growth from the melt of the novel material Dy3+:Lu2O3 and present new spectroscopic results for the mid-infrared spectral range. By slowly cooling down the melt in the crucible, we grew several high-quality samples of this material and performed a thorough spectroscopic analysis. The resulting cross-sections improve on the commonly used ZBLAN glass by a factor of four in absorption and more than a factor of two in emission, reaching 2.2 × 10−20 cm2 at a wavelength of 1260 nm and 6.7 × 10−21 cm2 at a wavelength of 2777 nm. Gain calculations suggest possible amplification in a wavelength range from 3.00 µm to 3.75 µm. Transmission and luminescence measurements at cryogenic temperatures enable the first determination of the energetic positions of the energy levels of this material up to an energy of 14,000 cm−1. Measurements of the decay dynamics were conducted and revealed a room temperature fluorescence lifetime of 50 µs for the 6H13/2 multiplet. Our results indicate Dy3+:Lu2O3 being a candidate for achieving mid-infrared lasing with this trivalent ion in an oxide host material.
2. Crystal growth
The starting materials Dy2O3 and Lu2O3 were mixed according to the desired doping concentrations (0.1 at.%, 0.5 at.%, 3.0 at.%) and placed in a crucible with a diameter of 17 mm and a height of 23 mm. Due to the high melting point of 2450 °C of Lu2O3 [19] a rhenium crucible was used. The setup was completed with an insulation of zirconia felts and alumina ceramic plates and held by a quartz glass tube (cf. Fig. 1(a)).
Fig. 1 Growth setup (a) and a grown boule of Dy3+(3 at.%):Lu2O3 (b). Despite the large surface area leading to the crystallization starting at different points, the boule contained large single crystalline volumes.
The rhenium crucible was inductively heated to the melting point of the growth material and, after subsequent cooling to room temperature, more growth material was added. This procedure was repeated until a sufficient volume of the crucible was filled with crystalline material (usually four to five iterations). The growth chamber was filled with a reducing atmosphere comprised of 95% nitrogen and 5% hydrogen as rhenium is prone to oxidization at higher temperatures. After the last crucible refill, the contents were melted again and the temperature was slowly lowered below the melting point over the course of 24 hours, leading to the formation of the final crystal boule. For in situ annealing 300 ppm of O2 were added to the atmosphere at a heating power corresponding to a growth temperature slightly below the melting point and the temperature was kept constant for a further three days. It was then lowered back down to room temperature over the course of three days, corresponding to a gradient of ~30 K/h, to prevent cracking of the grown crystal from thermal strain.
The obtained boules were about 12 mm in height, had a diameter of roughly 15 mm (cf. Fig. 1(b)) and typically consisted of few large single crystalline pieces. They were of very high quality and initial transmission experiments confirmed the absence of any visible scattering. A sample with dimensions of 7.2 mm × 2.6 mm × 1.9 mm used in the following experiments was cut and polished from one of these boules in our laboratories.
3. Experiments in detail
The room temperature transmission measurements utilized a Varian Cary 5000 dual beam spectrophotometer covering a wavelength range from 0.3 µm to 3.0 µm. Additional transmission measurements, as well as fluorescence measurements were conducted with a Horiba Jobin Yvon 1 m grating monochromator. The experimental setup is shown in Fig. 2. A broadband tungsten light source (ThorLabs SLS202L), providing a blackbody spectrum at 1900 K, is focused onto the sample. The transmitted light is collected and sent through the grating spectrometer into the detector unit which, depending on the detected wavelength range, consisted either of an Si, InGaAs, or InSb detector to cover the wavelength range from 0.7 µm to 4 µm. By using a chopper and a lock-in amplifier the signal-to-noise ratio of the measurement is improved before recording the data. A measurement without a sample was used for calibration.
Fig. 2 Experimental spectroscopy setups used for transmission and fluorescence measurements. For the fluorescence measurements the sample was slightly tilted and an excitation source was focused on it under grazing incidence. In case of the cryogenic measurements at 11 K a cryostat head was put into the sample position.
The fluorescence measurements are conducted in the same way except that an excitation laser beam, originating from a custom-built Er3+:Lu2O3 laser operating at 2845 nm, is focused onto the sample at grazing incidence in order to avoid reabsorption. Even though this laser is not resonant with one of the pronounced absorption lines of Dy3+:Lu2O3 (cf. Fig. 6(b)), the absorption was still high enough to conduct fluorescence measurements. Neither the sample aperture nor the broadband tungsten light source are used during fluorescence measurements. The chopper is relocated to the excitation beam. In a separate measurement the fluorescence detection unit was calibrated with the SLS202L blackbody radiation source. The fluorescence was measured at a wavelength of 3.006 µm.
Due to the Kramers-degeneracy in Dy3+ each multiplet – denoted by the term symbol 2S + 1LJ – is expected to contain (2J + 1)/2 Stark levels (cf. Fig. 3). In order to determine the position of the Stark levels of all the multiplets up to 6F3/2 the sample was mounted in a closed cycle helium cryostat. After cooling to around 11 K, transmission measurements were conducted using a Si detector head. The absorption into the multiplets 6F3/2, 6F5/2, 6F7/2, 6H5/2, 6F9/2, and 6H7/2 could thus be determined in the wavelength range from 725 nm to 1100 nm. The grating used for these measurements was blazed for 1 µm and featured a groove density of 600 mm−1. The resolution in this case was 0.075 nm.
Fig. 3 Energy level diagram for Dy3+:Lu2O3 with corresponding absorption and emission transitions of interest. In addition, possible pump and laser ESA as well as a possible cross-relaxation process are listed. Room temperature absorption spectra are aligned on the right and give an indication to the absorption strength for each multiplet.
The transmission measurements for determining the absorption into the 6F11/2, 6H9/2, and 6H11/2 multiplets were conducted with an InGaAs detector and a grating blazed for 1.5 µm at 600 grooves/mm. The slit width was set to a resolution of 0.3 nm.
The absorption into the 6H13/2 multiplet, as well as the emission from the same multiplet into the ground state 6H15/2 are both located in the mid-infrared spectral range around 3 µm. The corresponding transmission and fluorescence measurements utilized an LN2-cooled InSb detector. The resolution for these mid-infrared measurements was 0.3 nm for absorption and had to be increased to 7.2 nm for emission to yield a sufficient signal-to-noise ratio. Lifetime measurements were performed with the lock-in technique after Edmondson et al. [20] utilizing the excitation setup shown in Fig. 2.
4. Results and discussion
The cryogenic absorption spectra obtained from the transmission measurements at 11 K are displayed in Fig. 4. The grey vertical lines mark the determined positions of the Stark levels. The corresponding numerical data can be found in Table 1. Entries in the table marked with an asterisk (*) designate uncertain positions where an accurate determination could not be performed.
Fig. 4 Absorption cross section spectra for Dy3+(3 at.%):Lu2O3 at 11 K. The grey vertical lines mark the determined positions of the Stark levels. In the blue graph the labels for the energy multiplets are always to the right of the corresponding set. The remaining graphs only cover one multiplet each.
Table 1. Determined Stark energy level positions for Dy3+:Lu2O3 at 11 K.
The unmarked peaks in the mid-infrared cryogenic absorption spectrum, ranging from 3500 cm−1 to 4100 cm−1 are caused by absorption starting from the second Stark level of the ground state multiplet 6H15/2. Due to the small energetic separation from the ground state, this level still possesses a thermal population of 0.02% at 11 K and it can be assumed that the absorption in the sample leads to a local heating even increasing this thermal population. The small side peak at 3590 cm−1 shifted by 8 cm−1 from the zero phonon line could be caused by a weak transition of Dy3+ at the C3i site in the bixbyite structure of Lu2O3. The usual rare earth spectra originate from the acentric C2 site of the structure. The C3i site with its inversion symmetry strongly suppresses the transition probabilities, which could as well lead to additional weak absorption lines. Another origin of the side peak could be Davydov splitting [21] due to dimer formation involving two neighboring Dy3+ ions. The energy difference between the peaks corresponds to what would be attainable with this effect. Phonon interaction was ruled out by THz absorption measurements which revealed a minimum phonon energy in Lu2O3 of 90 cm−1 [22].
The emission from the 6H13/2 multiplet gives an insight into the positions of the Stark levels of the 6H15/2 ground state and is shown, calibrated to the zero-phonon line, in Fig. 5. For this ground state, the two uppermost Stark levels could not be clearly identified.
Fig. 5 Cryogenic emission spectrum of Dy3+:Lu2O3 calibrated to its zero-phonon line. The saturation near 100 cm−1 is caused by the intensity of the excitation source. The grey vertical lines mark the determined positions of the Stark levels.
The data gathered this way are shown in Table 1 and the corresponding energy level diagram is displayed in Fig. 3.
From room temperature transmission measurements the absorption cross-sections were calculated by the Beer-Lambert law (cf. Fig. 6). At a wavelength of 1.26 µm the absorption cross-sections display a maximum of 2.2 × 10−20 cm2, which is almost a factor of four higher compared to Dy3+:ZBLAN [23]. At a wavelength of about 2.75 µm the absorption cross-sections of Dy3+:Lu2O3 are still a factor of two higher compared to Dy3+:ZBLAN [24].
Fig. 6 Absorption (a) and emission (b) cross-sections of Dy3+:Lu2O3, with a resolution of 1 nm and 7.2 nm respectively.
Furthermore, multiple absorption peaks around a wavelength of 2.8 µm present several opportunities to realize a resonant pumping scheme. A possible pump source would be an Er3+-doped laser operating at or near a wavelength of 2.713 µm or 2.777 µm.
The emission cross-sections were calculated from the room temperature absorption spectra, via the McCumber [25] relation
where σem and σabs are the emission and absorption cross-sections, respectively, λ is the wavelength, Zl and Zu denote the partition functions for the lower and upper multiplet, respectively, and Ezpl is the energy of the zero-phonon line transition, in combination with the Füchtbauer-Ladenburg [26,27] equationwhere Ifl corresponds to the measured fluorescence intensity, τrad to the radiative lifetime, n to the refractive index of the material, and βu→l to the transition branching ratio.The McCumber relation yields reliable values for the emission cross sections in spectral regions where strong absorption is present, while it results in a weak signal-to-noise ratio in the longer wavelength range where only weak absorption is present. On the other hand, the measured fluorescence and thus the outcome of the Füchtbauer-Ladenburg equation is strongly influenced by reabsorption in the short wavelength range, but delivers reliable values for the long wavelength range. Therefore, the emission cross section spectra up to a wavelength of 2.895 µm were calculated by Eq. (1) and the outcome of Eq. (2) was matched to these values in the range between 2.805 µm and 2.972 µm yielding a continuous emission cross-section spectrum in the range between 2.250 µm and 3.800 µm. This procedure also enabled an estimation of the radiative lifetime of the emitting multiplet 6H13/2 by Eq. (2) to 20 ms.
The Judd-Ofelt theory allows to directly calculate the radiative lifetime from the absorption cross sections [28], yielding 28 ms, which is in reasonable agreement with the previously explained method.
The maximum emission cross-sections were determined as 6.7 × 10−21 cm2, which is a factor of two higher than the maximum in Dy3+:ZBLAN. At longer wavelengths this factor becomes even larger with up to a sixfold increase over Dy3+:ZBLAN at a wavelength of 3.246 µm [29], clearly showing the potential of Dy3+:Lu2O3 for lasing in this wavelength range.
In the gain cross-section spectra calculated according to
a first onset of gain is found at an inversion level β of around 0.03 in the wavelength range between 3.385 µm and 3.8 µm. For higher inversions the maximum gain shifts towards shorter wavelengths and for inversion levels above 0.1 we expect an onset of laser action at 3.246 µm (cf. Fig. 7). At inversion levels exceeding 0.2 the expected gain range extends from 3.0 µm until the end of the measurement range at 3.8 µm. Compared to the recently published tunable Dy3+:ZBLAN laser by Majewski et al. [29], who realized a tuning range from 2.8 µm to 3.4 µm, this would enable access to even longer wavelengths.However, investigation of the fluorescence dynamics showed a fluorescence lifetime for the 6H13/2 multiplet of 50 µs for the 3 at.%-doped sample as well as 70 µs for the 0.5 at.%-doped sample. This is three orders of magnitude shorter than the estimated radiative lifetime of 20 ms to 28 ms. The reason for the large discrepancy between the estimated radiative and the measured fluorescence lifetime is an increased rate of non-radiative transitions due to the small energy gap between the 6H13/2 and the 6H15/2 multiplets of 2743 cm−1, which corresponds to a wavelength of 3.645 µm. This is equal to roughly 4.4 times the maximum phonon energy of Lu2O3 of 613 cm−1. According to research conducted by Moos [30], significant non-radiative decay can be expected for transitions which bridge an energy gap of less than five times the maximum phonon energy.
Measurements regarding pump ESA did not yield any usable output signal when using the ESA measurement setup described in [31]. The lack of matching excitation sources and the corresponding off-peak excitation led to a too low inversion density and insufficient signal-to-noise ratio. On the other hand, pump ESA, as mentioned before, is only expected to occur – if at all – for short pump wavelengths. Due to the lack of a useful, wavelength-matched pump source it was also not possible to perform meaningful laser experiments.
5. Conclusion
The cubic sesquioxide Dy3+:Lu2O3 has, to the best of our knowledge, for the first time been grown from the melt. The cooling down method yielded crystal samples of excellent quality. The energetic positions of the Stark energy levels of this material were determined for the first time up to an energy of 14,000 cm−1. Subsequent spectroscopic analyses revealed up to four times higher absorption and six times higher emission cross-sections compared to Dy3+:ZBLAN. Broad gain should be available over a wavelength range of ~700 nm around 3.4 µm.
Considering laser operation on the ground state transition 6H13/2 → 6H15/2 and pumping at wavelengths of 1.25 µm (→ 6F11/2), the cross relaxation process 6F11/2, 6H15/2 → 6H13/2, 6H13/2 depicted in Fig. 3 is nearly resonant and could thus contribute to a 2-for-1 pumping mechanism similar to the process observed in Tm3+-lasers at 2 µm. As the energy gaps between the multiplets above the 6H13/2 multiplet are small, the decay processes are dominated by non-radiative transitions. This would also hold for pumping into all levels above the 6F11/2 multiplet.
The energy gap between the ground state and the 6H13/2 multiplet of 2743 cm−1, equal to 4.4 times the maximum phonon energy of Lu2O3 of 618 cm−1 [11], should be large enough to allow for a laser transition to take place. Instead of relying on a possible cross-relaxation process and pumping at shorter wavelengths there exists also the possibility of resonant pumping directly into the 6H13/2 multiplet at wavelengths around 2.8 µm.
While the increased rate of non-radiative transitions is expected to increase the laser threshold and temperature of the active volume, it should not prevent this material from showing laser action in the mid-infrared spectral range – at least under wavelength-matched pulsed pumping. As the short fluorescence lifetime will make cw lasing at room temperature difficult, a cryogenic approach could be leveraged in order to achieve mid-infrared lasing with this material. At cryogenic temperatures the transitions due to phonon interactions should be suppressed which should increase the fluorescence lifetime of the 6H13/2 multiplet. In addition, using a matching pump source could also allow for the determination of the fluorescence lifetime of samples with even lower doping concentrations and improve the absorption efficiency.
From the results of our research, we regard Dy3+:Lu2O3 as a promising candidate for oxide-based laser operation in the mid-infrared spectral range, providing access to a wavelength range between 3.0 µm and 3.75 µm.
Funding
BMBF project “TSUNAMI” (FKZ 13N13050) and “EQuiLa” (FKZ 13N14192); excellence cluster ‘The Hamburg Centre for Ultrafast Imaging - Structure, Dynamics and Control of Matter at the Atomic Scale’ of the Deutsche Forschungsgemeinschaft.
References
1. J. Mulrooney, J. Clifford, C. Fitzpatrick, and E. Lewis, “Detection of carbon dioxide emissions from a diesel engine using a mid-infrared optical fibre based sensor,” Sens. Actuators A Phys. 136(1), 104–110 (2007). [CrossRef]
2. S. J. Linke, A. Frings, L. Ren, A. Gomolka, U. Schumacher, R. Reimer, N.-O. Hansen, N. Jowett, G. Richard, and R. J. D. Miller, “A new technology for applanation free corneal trephination: The picosecond infrared laser (PIRL),” PLoS One 10(3), e0120944 (2015). [CrossRef] [PubMed]
3. X. C. Zhang, A. Shkurinov, and Y. Zhang, “Extreme terahertz science,” Nat. Photonics 11(1), 16 (2017). [CrossRef]
4. R. R. Gattass, R. Thapa, F. H. Kung, L. E. Busse, L. B. Shaw, and J. S. Sanghera, “Review of infrared fiber-based components,” Appl. Opt. 54(31), F25–F34 (2015). [CrossRef] [PubMed]
5. R. I. Woodward, M. R. Majewski, G. Bharathan, D. D. Hudson, A. Fuerbach, and S. D. Jackson, “Watt-level dysprosium fiber laser at 3.15 μm with 73% slope efficiency,” Opt. Lett. 43(7), 1471–1474 (2018). [CrossRef] [PubMed]
6. M. R. Majewski and S. D. Jackson, “Highly efficient mid-infrared dysprosium fiber laser,” Opt. Lett. 41(10), 2173–2176 (2016). [CrossRef] [PubMed]
7. S. D. Jackson, “Continuous wave 2.9 µm dysprosium-doped fluoride fiber laser,” Appl. Phys. Lett. 83(7), 1316–1318 (2003). [CrossRef]
8. M. Bass, Handbook of Optics (McGraw-Hill, 1995).
9. A. Kaminskii, Laser Crystals: Their Physics and Properties, in Springer Series in Optical Sciences (Springer Berlin Heidelberg, 2013).
10. R. Peters, C. Kränkel, S. Fredrich-Thornton, K. Beil, K. Petermann, G. Huber, O. Heckl, C. Baer, C. Saraceno, T. Südmeyer, and U. Keller, “Thermal analysis and efficient high power continuous-wave and mode-locked thin disk laser operation of Yb-doped sesquioxides,” Appl. Phys. B 102(3), 509–514 (2011). [CrossRef]
11. E. Mix, “Kristallzüchtung, Spektroskopie und Lasereigenschaften Yb-dotierter Sesquioxide,” Ph.D. thesis, Universität Hamburg (1999).
12. L. Laversenne, Y. Guyot, C. Goutaudier, M. T. Cohen-Adad, and G. Boulon, “Optimization of spectroscopic properties of Yb3+-doped refractory sesquioxides: cubic Y2O3, Lu2O3, and monoclinic Gd2O3,” Opt. Mater. 16(4), 475–483 (2001). [CrossRef]
13. A. M. Heuer, C. J. Saraceno, K. Beil, G. Huber, and C. Kränkel, “Efficient OPSL-pumped mode-locked Yb:Lu2O3 laser with 67% optical-to-optical efficiency,” Sci. Rep. 6(1), 19090 (2016). [CrossRef] [PubMed]
14. P. von Brunn, A. M. Heuer, L. Fornasiero, G. Huber, and C. Kränkel, “Efficient laser operation of Nd3+:Lu2O3 at various wavelengths between 917 nm and 1463 nm,” Laser Phys. 26(8), 084003 (2016). [CrossRef]
15. T. Li, K. Beil, C. Kränkel, and G. Huber, “Efficient high-power continuous wave Er:Lu2O3 laser at 2.85 μm,” Opt. Lett. 37(13), 2568–2570 (2012). [CrossRef] [PubMed]
16. M. Guzik, J. Pejchal, A. Yoshikawa, A. Ito, T. Goto, M. Siczek, T. Lis, and G. Boulon, “Structural Investigations of Lu2O3 as Single Crystal and Polycrystalline Transparent Ceramic,” Cryst. Growth Des. 14(7), 3327–3334 (2014). [CrossRef]
17. J. Shi, B. Liu, Q. Wang, H. Tang, F. Wu, D. Li, H. Zhao, Z. Wang, W. Deng, X. Xu, and J. Xu, “Crystal growth, spectroscopic characteristics, and Judd-Ofelt analysis of Dy:Lu2O3 for yellow laser,” Chin. Phys. B 27(7), 077802 (2018). [CrossRef]
18. A. M. Heuer, P. von Brunn, and C. Kränkel, “Spectroscopy of Er3+-, Dy3+-, Pr3+-, and Ho3+-doped sesquioxides in the mid-infrared spectral region”, 7th EPS-QEOD Europhoton Conference, Vienna, Austria, p33.22 (2016).
19. R. Peters, K. Petermann, and G. Huber, “Growth technology and laser properties of Yb-doped sesquioxides, in Crystal Growth Technology - Semiconductors and Dielectrics, P. Capper and P. Rudolph, eds. (Wiley, 2010), pp. 267–282.
20. K. Edmondson, S. Agoston, and R. Ranganathan, “Impurity level lifetime measurements using a lock-in amplifier,” Am. J. Phys. 64(6), 787–791 (1996). [CrossRef]
21. H. B. Ergun, K. A. Gehring, and G. A. Gehring, “Jahn-Teller induced Davydov splitting in TbVO4,” J. Phys. Chem. 9, 1101 (1976).
22. A. Toncelli, Dipartimento di Fisica, Universitá di Pisa, Largo Bruno Pontecorvo 3, Pisa, Italy, (personal communication, 2018).
23. Y. H. Tsang, A. E. El-Taher, T. A. King, and S. D. Jackson, “Efficient 2.96 µm dysprosium-doped fluoride fibre laser pumped with a Nd:YAG laser operating at 1.3 µm,” Opt. Express 14(2), 678–685 (2006). [CrossRef] [PubMed]
24. Y. H. Tsang and A. E. El-Taher, “Efficient lasing at near 3 µm by a Dy-doped ZBLAN fiber laser pumped at 1.1 µm by an Yb fiber laser,” Laser Phys. Lett. 8(11), 818–822 (2011). [CrossRef]
25. D. E. McCumber, “Einstein relations connecting broadband emission and absorption spectra,” Phys. Rev. 136(4A), A954–A957 (1964). [CrossRef]
26. C. Füchtbauer, G. Joos, and O. Dinkelacker, “Über Intensität, Verbreiterung und Druckverschiebung von Spektrallinien, insbesondere der Absorptionslinie 2537 des Quecksilbers,” Ann. Phys. 376(9-12), 204–227 (1923). [CrossRef]
27. R. Ladenburg, “Die quantentheoretische Deutung der Zahl der Dispersionselektronen,” Z. Phys. 4(4), 451–468 (1921). [CrossRef]
28. B. Henderson and R. H. Bartram, Crystal-Field Engineering of Solid-State Laser Materials, Cambridge Studies in Modern Optics (Cambridge University Press, 2005).
29. M. R. Majewski, R. I. Woodward, and S. D. Jackson, “Dysprosium-doped ZBLAN fiber laser tunable from 2.8 µm to 3.4 µm, pumped at 1.7 µm,” Opt. Lett. 43(5), 971–974 (2018). [CrossRef] [PubMed]
30. H. W. Moos, “Spectroscopic relaxation processes of rare earth ions in crystals,” J. Lumin. 1, 106–121 (1970). [CrossRef]
31. J. Koetke and G. Huber, “Infrared excited-state absorption and stimulated emission cross sections of Er3+-doped crystals,” Appl. Phys. B 61(2), 151–158 (1995). [CrossRef]
