A fine-tunable white light source using a novel dual-excitation design and efficient Sn-doped condensed borate glass phosphor

Abstract

Efficient and promising luminescent borate glass doped with tin was fabricated and examined. This unique optical material is characterized by two considerably different Sn emission centers affected by distinct crystal field strengths. Firstly, the efficient broadband luminescence can be selectively excited at 266 nm and 360 nm; however, simultaneous activation of both blue and red emission bands is possible as a result of the overlapping of PLE spectra. Consequently, intense and broad emission covering the whole visible light spectrum with an extremely high CRI of 97 is achieved. This result is exceptional for a single phosphor light source. The resulting luminescence glass color can be smoothly adjusted from cool-white to warm-white utilizing anti-overlapping or overlapping excitation ranges especially between 300 and 340 nm. A dual-wavelength-excitation experiment was performed to demonstrate the excellent control of the balance between the blue and red emissions by varying the power of the excitation diode. The highest quantum yield (QY) of the LaMBO:Sn emission was found to be 85%.

---

Introduction

White or specific color light-emitting sources based on amorphous or crystalline materials doped with optically active centers are intensively examined owing to their satisfactory quantum yield of emission, high color rendering index, considerable brightness and low energy consumption.^1,2 Borate glasses have been recognized as advanced optical materials owing to B₂O₃ superstructures consisting of planar BO₃ and tetrahedral BO₄, which may interact to form numerous borate structural units depending on the number of non-bridging oxygens (NBOs) or bridging oxygens (BOs).^3,4 The maximal phonon energy of borates is quite high; however, it can be significantly reduced exemplarily by incorporation of alkali and alkaline metal ions into the oxide glasses.⁵ Moreover, borate glasses are characterized by a quite low melting point, high thermal stability and good solubility of luminescent ions.⁶ Consequently, various admixtures are utilized to reveal wide-range luminescence in borate glasses and these amorphous materials have been found to be an appropriate alternative to resin based phosphors.⁷ It is crucial since the effective light sources can be designed free from organic resin having poor thermal resistance.⁸ Quite recently, Wang et al. examined a copper-containing borate glass as a potential optical storage and encryption medium. The UV-excited broadband emission is reversibly regulated by alternating 360 nm laser irradiation and thermal treatment at 400 °C.⁹ The potential of dysprosium doped borate–zinc–lithium–aluminum glasses for producing efficient visible luminescence was examined and the CIE coordinates implied that BZLADy glass emission is located within the white light region.¹⁰ The useful broadband Eu²⁺ luminescence related to 4f⁶5d¹ → 4f⁷ transitions was explored in barium–aluminum–borate oxyfluoride glasses and additionally, the 4f⁷ → 4f⁷ transition of Eu²⁺ appeared in the glass ceramics after heat treatment.¹¹

The optical qualities of luminescent Sn²⁺ ions doped in distinct inorganic hosts are highly advantageous as well. Some authors have considered the incorporation of Sn into the structure of glasses for relevant bio-medical applications.¹² The study of this issue has been ongoing since the 1960s.¹³ Energy transfer from Sn²⁺ to Mn²⁺ ions in single crystals of KBr co-doped with tin and manganese ions was discovered thirty years later.¹⁴ The achieved data for aluminophosphate glass implied that light absorption might take place at two-coordinated tin centers, followed by energy transfer to dysprosium ions.¹⁵ The luminescence properties of Sn-doped and Er-co-doped SiO₂ glasses were investigated by Paleari et al.¹⁶ The quenching process of Sn-doped sol–gel silica luminescence was examined in a wide temperature range of 18–300 K.¹⁷ A donor/acceptor energy transfer was observed in Sn,Sm-co-doped P₂O₅:BaO phosphate glass and a broad blue-white band characteristic of two-coordinated Sn centers and orange-red emission bands of Sm³⁺ ions were attained.¹⁸ The optical properties of Cu⁺/Sn²⁺, Dy³⁺/Sn²⁺, Er³⁺/Sn²⁺, Eu³⁺/Sn²⁺ or Yb³⁺/Sn²⁺ co-doped phosphate glasses were assessed by Jiménez et al.^19–23 The enhancement of UV gadolinium luminescence was observed in Sn/Gd co-doped Li₂O-PbO-P₂O₅ glass.²⁴ The coexistence of fluorescence and phosphorescence was discovered in tin fluorophosphate glass.²⁵ Cool white luminescence composed of Sn²⁺ blue-green broadband emission and red emission of Eu³⁺ was detected in P₂O₅-SnO₅-SnF₂ glass.²⁶ In contrast to amorphous phosphate materials, studies on Sn-doped borate glasses are relatively scarce. The optical, radioluminescence and thermoluminescence (TL) properties of Sn-doped magnesium aluminoborate glasses were examined by Masai et al. It was found that tin-doped xSnO-(60 − x)ZnO–40B₂O₃ borate glasses reveal thermally stimulated X-ray-induced luminescence.²⁷ Moreover, broadband tunable blue-violet emissions ranging from 320 to 550 nm were realized in Sn²⁺ doped boroaluminate glasses.²⁸

In this paper, we report a perspective borate glass phosphor singly doped with optically active tin ions. The extremely high color rendering index (97) and highest quantum yield (QY, 85%) may make the material under study an efficient component for advanced lighting devices. For this phosphor, the effective broadband luminescence ranging from cool-white to warm-white light can be effectively tuned and controlled utilizing alternatively the excitation wavelengths and/or the power of the excitation source. The continuous tunability of the dual-excited luminescence is demonstrated to obtain the correlated color temperature (CCT), especially in an excellent wide range between 4500 and 7500 K.

Experimental

Glass preparation

The sample with the composition 95 (62.5 B₂O₃–12.5 La₂O₃–25 MgO)–5 SnO was prepared using the following starting materials: La₂O₃ (99.99%, Alfa Aesar), MgO (pure p.a., Reachim), H₃BO₃ (pure p.a., POCH) and SnO (99.9%, Alfa Aesar). The materials were weighed, mixed and ground in a porcelain mortar and melted in a platinum crucible at 1300 °C in an air environment. After 10 minutes, upon reaching the melting temperature, the melt was removed from the furnace and quenched onto a room temperature brass plate. Subsequently, the glass sample was annealed at 450 °C for 12 h to reduce residual stress.

Measurement and characterization

Structural characterization by X-ray diffraction (XRD) has been performed using an X'Pert PRO powder diffractometer (PANalytical) equipped with a linear PIXcel detector and Cu Kα radiation (λ = 1.54056 Å). The mid-IR transmittance spectra were recorded using a Nicolet iS50 FTIR spectrometer using KBr pellets. The spectra were recorded with a resolution of 2 cm⁻¹. The absorption spectrum was recorded using a Varian Cary 5E (Agilent) UV-vis-NIR spectrophotometer. Emission spectra were recorded using a Hamamatsu PMA-12 spectrophotometer equipped with a BT-CCD linear image sensor (Hamamatsu Photonics K.K, Shizuoka, Japan) and 266 nm, 360 nm and occasionally 375 nm laser diodes as excitation sources. The temperature of the sample was controlled using a Linkam THMS 600 heating/freezing stage (The McCrone Group, Westmont, IL, USA). The power of the excitation source was controlled using the applied current (360 nm laser diode) or by using neutral density filters (266 nm laser diode). The PLE and decay time measurements were performed using an FLS-1000 spectrophotometer (Edinburgh Instruments, Edinburgh, UK) in the Czerny–Turner configuration equipped with a micro-second pulse lamp and 450 W xenon lamp. The quantum yield measurements were recorded using a Hamamatsu PMA-12 spectrophotometer, with the sample placed in an integrating sphere. The photographs were captured using a 256 nm/365 nm UV lamp and an iPhone SE2 camera. The video was recorded using 266 nm and 360 nm laser diodes and an iPhone SE2 camera.

Results and discussion

The resulting glass is colorless and transparent (Fig. 1a). The dopant causes no apparent coloration, which can also be verified using the absorption spectrum, where only the absorption edge at ∼300 nm is present (Fig. 1c). The IR spectrum of the LaMBO:Sn glass (Fig. 1c) is consistent with previous reports on LaMBO glass²⁹ and consists of three main bands located in 500–750 cm⁻¹, 750–1200 cm⁻¹ and 1200–1600 cm⁻¹ regions associated with the bending vibrations of B–O–B linkages, the B–O stretching vibrations of BO₄ groups and the B–O stretching vibrations of BO₃ groups, respectively.³⁰ The lack of any crystal phase is confirmed by the XRD results, in which only a broad scattered signal is observed without sharply defined diffraction lines (Fig. 1d).

Fig. 1 (a) Photograph, (b) absorption, (c) IR spectrum and (d) XRD results of LaMBO:Sn glass.

The luminescence properties of LaMBO:Sn strongly indicate the presence of two distinct emission centers. The selective excitation at 266 nm and 365 nm allows observing two distinct broad emission bands with maxima at 420 nm and 650 nm (Fig. 2a), which will be later on labeled the “blue band” and “red band”, respectively. The excitation spectra recorded by selective monitoring of blue and red bands revealed that the former is excited most efficiently at 260 nm, while the latter is excited most efficiently at 295 nm. Both blue and red emissions are characterized by a large Stokes shift: the blue emission has a Stokes shift value equal to 15 035 cm⁻¹, while the red emission has a Stokes shift value of 18 640 cm⁻¹ (Fig. S1†).

Fig. 2 (a) Excitation (PLE) and emission (PL) spectra of LaMBO:Sn glass. (b) Photographs of LaMBO:Sn glass under 256 nm and 365 nm UV lamp illumination. (c) Luminescence decay curves of cool white and warm white emissions.

The two-band luminescence of LaMBO:Sn is visible to the naked eye under UV excitation at different wavelengths (Fig. 2b). Although the Sn²⁺ ions’ excitation spectrum can be decomposed into two bands resulting from transitions ¹S₀ → ³P₁ (the so-called A-band) and ¹S₀ → ¹P₁ (the so-called C-band), these excitation bands do not account for the dual-emission of LaMBO:Sn, since both mentioned transitions result in the same emission spectrum.³¹ Another explanation for the observed phenomenon is the triplet–singlet and singlet–singlet transitions, described by Skuja³² for two-fold coordinated Sn²⁺ ions in glassy silica, but the singlet–singlet emission was recorded in the 270–340 nm range and its Stokes shift is sub-10 000 cm⁻¹, which is not consistent with our results for the LaMBO:Sn optical system. This effective dual emission can be considered based on an idea given by Smets³³ and van der Steen,³⁴ who postulate a perturbation of the ³P₁ level by the Jann–Teller effect, leading to the occurrence of two minima in the adiabatic potential energy surface of the relaxed excited state. A more probable explanation of the dual emission in LaMBO:Sn is the analogous luminescence in Sn-doped KBr³⁵ and KI,³⁶ where the red emission is associated with Sn²⁺–Sn²⁺ dimers. A similar observation for trivalent bismuth in silica glasses leads to the conclusion that the additional red luminescence can be produced either by dimers or complexes consisting of a dopant and an oxygen vacancy.³⁷

Another possibility is the presence of Sn⁴⁺ ions resulting from the oxidation process during sample preparation at high temperature in air. Sn⁴⁺ ions were reported to emit red light under blue light excitation.^38,39

To further study the two-band luminescence phenomenon, decay curves of blue and red bands were recorded and fitted with a bi-exponential equation (eqn (1)). In Sn²⁺, the bi-exponential nature of the decay results from the repopulation of the emitting ³P₁ level by the ³P₀ trap level.³¹

where I is the luminescence intensity, A and B are the intensity factors of decay components, t is the time, and τ₁ and τ₂ are the decay time constants. The blue luminescence is characterized by shorter decay constants equal to τ₁ = 5.4 μs and τ₂ = 10.6 μs, while the decay constants of the red luminescence are longer: τ₁ = 26.6 μs and τ₂ = 78.0 μs. This result indicates that both bands result from the partly forbidden transitions, characterized by microsecond decay times.³² The result is also consistent with the blue emission originating from Sn²⁺ ions and the red emission originating from Sn⁴⁺ ions, for which the decay time was reported to be 120 μs in phosphate glass.³⁸ It is nevertheless inconclusive.

Temperature-dependent emission spectra were recorded to investigate the thermal stability of blue and red band emissions of LaMBO:Sn. The blue emission spectra exhibit additional bands at short wavelengths (360–400 nm range) resulting from the trap ³P₀ level.³⁴ These additional bands disappear quickly with increasing temperature and are absent for temperatures above −150 °C (Fig. 3a). The integral intensity of the blue emission is stable in the temperature range from −193 to −75 °C and decreases steadily for higher temperatures, due to conventional thermal quenching processes (Fig. 3b). Fitting with the Arrhenius equation reveals a single thermal depopulation pathway with an activation energy of 1054 cm⁻¹ (Fig. S2†). The material was assessed with respect to its utility as a lighting phosphor by performing heating and cooling cycles between room temperature and the standard chip LED working temperature. The repeatability of its integral intensity was calculated using eqn (2).

where I^{25 °C}_av is the average integral intensity at 25 °C and max(I_av − I) is the maximum deviation of the integrated intensity value from the average value at a given temperature. The repeatability of the blue emission in the 20–100 °C cycle is 98.1% (Fig. 3c).

Fig. 3 (a) Temperature-dependent emission spectra of blue emission under 266 nm excitation. (b) PL intensity thermal quenching curve of blue emission. (c) Repeatability of the blue band PL intensity during the heating–cooling cycle of 25–100 °C.

An analogous assessment was performed for the red band of LaMBO:Sn emission (Fig. 4a). The red band is significantly less thermally stable than the blue band and its intensity drops as the temperature increases in the whole examined range (Fig. 4b). The poorer stability is a result of lower activation energy of thermal quenching. This is consistent with the observation of Scacco et al.³⁵ and contrasts withthe result reported by Skuja,³² which further indicates the dimer explanation of the red emission to be accurate. The activation energy was determined from eqn (S1) and (S2) in the ESI† to be 240 cm⁻¹ (Fig. S2†). The red band maximum undergoes a blue shift with the temperature increase from 683 nm for −193 °C to 554 nm for 300 °C (Fig. S3†), which was also observed by Smets et al. for the lower-energy emission band in Ba₆Y₂AI₄O_I5:Sn crystals.³³ The repeatability of the red emission integrated intensity in the 25–100 °C heating–cooling cycle was determined analogically to the blue one using eqn (2) and is equal to 99.0% (Fig. 4c).

Fig. 4 (a) Temperature-dependent emission spectra of red emission under 266 nm excitation. (b) PL intensity thermal quenching curve of red emission. (c) Repeatability of the red band PL intensity during the heating–cooling cycle of 25–100 °C.

An additional temperature treatment was applied to further investigate the behavior of the red emission. As LaMBO:Sn was exposed to 500 °C for 4 hours, the red emission was not affected within the error range, but the heat treatment at 850 °C totally quenched the red emission after 15 minutes (Fig. S4†). These observations serve as final evidence that supports the previous supposition that the red emission is produced by a Sn²⁺–oxygen vacancy complex. Consequently, heat treatment at a sufficiently high temperature gives rise to a reduction in the number of such structural defects and the broadband red luminescence in LaMBO:Sn is effectively quenched.

Overlapping PLE spectra (Fig. 2a) indicate the possibility of activation of both blue and red emission bands by excitation in the 260–400 nm range. The shift of the excitation light influences the shape of the resulting emission spectrum, by modifying the ratio between the blue and red bands (Fig. 5a). While the excitation in the 260–290 nm and 350–400 nm ranges, which correspond to the anti-overlapping parts of the PLE spectra, results in practically one-band emission of comparable intensity, the excitation in the overlapping range of 300–340 nm results in very wide emission spanning the whole visible light spectrum. This phenomenon can be visualized on the CIE color coordinate triangle (Fig. 5b), where the color coordinates as a function of the excitation wavelength span the range from blue-white to cool white, neutral white and warm white, crossing the middle of the triangle. In the mentioned PLE overlapping range of 300–340 nm, the color rendering index (CRI) is extremely high and equal to 96–97, which is unusual for a single phosphor light source. Commercially available sources based on the single phosphor Ce:YAG coupled with a blue LED exhibit CRI values in the range of 62–83.⁴⁰

Fig. 5 (a) Emission spectra at different excitation wavelengths. (b) CIE color coordinates of the LaMBO:Sn emission at different excitation wavelengths and the corresponding CRIs. (c) Quantum yield of the LaMBO:Sn emission at different excitation wavelengths.

The quantum yield (QY) of the LaMBO:Sn emission is the highest for excitation at 260 nm and the internal QY is equal to 85% (Fig. 5c), making the blue emission much more efficient than the red one. The excitation in the PLE overlapping range of 300–340 nm, which produces white emission, has an internal QY in the range of 17–48%, which is lower than that of the conventional Ce:YAG phosphor (∼80%).⁴¹ The internal QY for the red emission under 360–400 nm excitation is below 12%. The external QY, which takes into account the absorbance factor, follows the same trend as the internal QY and has a maximum value of 47%.

From the practical point of view, it is easier to control the power of the two separate diodes than shifting the wavelength of the single diode. With this in mind, the dual-excitation PL spectra were recorded using two diodes with output wavelengths of 266 nm and 360 nm. The power of the latter was controlled continuously in the 2.2–14.3 mW range. Due to the imbalance in the quantum yield between excitations at those two wavelengths (Fig. 5c), the 266 nm diode must have a significantly lower power output (6 mW) than the 360 nm diode to obtain balanced white emission. The obtained spectra show that it is possible to control the balance between the blue and red emission bands by changing the power of the 360 nm diode (Fig. 6a). The resulting color of the emission is close to the ideal neutral white (Fig. 6b) and the correlated color temperature (CCT) can be continuously tuned between 4500 and 7500 K (Fig. 6c). The continuous tunability of the dual-excited emission is demonstrated in Video S1.† The obtainable color range can be shifted towards blue by increasing the power of the 266 nm diode to 10 mW, which allows obtaining a CCT of up to 11 000 K.

Fig. 6 (a) Emission spectra of LaMBO:Sn under dual excitation with 266 nm and 360 nm laser diodes. (b) CIE color coordinates of the LaMBO:Sn emission under dual excitation. (c) Correlated color temperature of the LaMBO:Sn emission under dual excitation with the corresponding color descriptions.

Conclusion

Sn-doped borate glasses were successfully manufactured by a conventional melt quenching method. Different tin emission centers may be recognized, since selective UV excitation results in two distinct broad emission bands with maxima at 420 nm and 650 nm. This finding is supported by a detailed study of the decay kinetics of luminescence. The LaMBO:Sn material was evaluated with respect to its utility as a lighting phosphor and consequently, the thermal stability of blue and red tin emissions was assessed. In particular, the repeatability of the luminescence in the 20–100 °C cycles exceeds 98%, indicating the unique qualities of the tin centers in the examined glass system. The ratio between the blue and red bands is significantly adjusted and the whole visible light spectrum is achieved when the overlapping 300–340 nm PLE excitation wavelengths are used. The dual-beam experiments showed that the resulting cold/warm white phosphor color and appropriate balance between the blue and red emissions are effectively modified and controlled, especially when the 360 nm diode power is adequately varied. In summary, LaMBO:Sn may be considered to be an excellent fluorescent glass host for advanced lighting devices since the related CRI (color rendering index) is 97 and the highest emission quantum yield (QY) of 85% is achieved with 260 nm excitation.

Author contributions

B.B.: conceptualization, data curation, formal analysis, investigation, methodology, visualization, writing – original draft, and writing – review and editing; R.L.: conceptualization, investigation, methodology, resources, writing – original draft, and writing – review and editing.

Data availability

Data will be made available on request.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors would like to acknowledge Maciej Ptak for performing the IR measurement and Ewa Bukowska for performing the XRD measurement. This research received no external funding.

References

1. R. Ding, Y. Chen, H. Wang, W. Li, Y. Su, Y. Sheng, X. An, L. Liu and C. Li, J. Lumin., 2024, 275, 120802.
2. R. Rajaramakrishna, P. Nijapai, P. Kidkhunthod, H. J. Kim, J. Kaewkhao and Y. Ruangtaweep, J. Alloys Compd., 2020, 813, 151914.
3. P. Deepa, P. Murugasen, A. A. Suresh, M. Dhavamurthy and A. V. Deepa, Emergent Mater., 2024, 7, 1715–1731.
4. D. Singh, K. Singh, B. S. Bajwa, G. S. Mudahar, D. P. Singh, Manupriya, M. Arora and V. K. Dangwal, J. Appl. Phys., 2008, 104, 103515.
5. M. Anastasopoulou, K. C. Vasilopoulos, D. Anagnostopoulos, I. Koutselas, D. K. Papayannis and M. A. Karakassides, J. Phys. Chem. B, 2017, 121, 4610–4619.
6. A. Madhu, A. S. Alqarni, U. K. Kagola and N. Srinatha, Appl. Phys. A, 2024, 130, 355.
7. S. Kaur, A. K. Vishwakarma, N. Deopa, A. Prasad, M. Jayasimhadri and A. S. Rao, Mater. Res. Bull., 2018, 104, 77–82.
8. F. Zhang, Q. Cao, J. Lu, H. Tao, S. Yang, S. Yao, Q. Fu, Z. Ma, W. Dai and H. Zhao, Ceram. Int., 2024, 50, 35557–35567.
9. A. Wang, A. Huang, H. Zhao, Y. Liu, Y. Luo, D. Zhou, Z. Song, J. Qiu, C. Tatiana and Z. Yang, Chem. Eng. J., 2024, 487, 150648.
10. N. Yaduvanshi, C. Basavapoornima, A. S. Alqarni, U. K. Kagola, T. Uthayakumar, A. Madhu and N. Srinatha, Opt. Mater., 2024, 151, 115359.
11. M. Zhao, P. Wang, C. Shao, W. Li, D. Chen and W. Chen, J. Mater. Chem. C, 2024, 12, 13411–13420.
12. U. Pantulap, M. Arango-Ospina and A. R. Boccaccini, J. Mater. Sci. Mater. Med., 2022, 33, 3.
13. E. Lendvay, Phys. Status Solidi B, 1962, 2, 460–472.
14. R. Aceves, U. G. Caldiño, O. J. Rubio and E. Camarillo, J. Lumin., 1995, 65, 113–119.
15. J. A. Jiménez, S. Lysenko and H. Liu, J. Mater. Sci., 2010, 45, 2983–2990.
16. A. Paleari, G. Brambilla, N. Chiodini, S. Ghidini and D. Di Martino, Photorefractivity and luminescence properties of Sn-doped SiO2 glasses, Proc. SPIE 4640, Integrated Optics: Devices, Materials, and Technologies VI, 2002, https://doi.org/10.1117/12.431874.
17. A. Cannizzo, S. Agnello, M. Cannas, N. Chiodini, M. Leone and A. Paleari, J. Non-Cryst. Solids, 2005, 351, 1937–1940.
18. J. A. Jiménez, J. Electron. Mater., 2014, 43, 3588–3592.
19. J. A. Jiménez, Opt. Mater., 2014, 37, 347–351.
20. J. A. Jiménez, J. Non-Cryst. Solids, 2014, 387, 124–127.
21. J. A. Jiménez, M. Sendova, B. Hosterman and L. Haney, Mater. Lett., 2014, 131, 344–346.
22. J. A. Jiménez and E. R. Fachini, J. Rare Earths, 2015, 33, 27–32.
23. J. A. Jiménez, D. Hayes and C. Hasselbrink, J. Phys. Chem. C, 2022, 126, 18421–18428.
24. Y. Gandhi, P. Rajanikanth, M. Sundara Rao, V. Ravi Kumar, N. Veeraiah and M. Piasecki, Opt. Mater., 2016, 57, 39–44.
25. W. Mao, W. Xie, P. Li, Y. Lu, Y. Duan, S. Xu and J. Zhang, Chem. Eng. J., 2020, 399, 125270.
26. W. Mao, M. Cai, W. Xie, P. Li, W. Zhen, S. Xu and J. Zhang, J. Alloys Compd., 2019, 805, 205–210.
27. H. Masai, H. Kimura, N. Kawaguchi, T. Yanagida and N. Kitamura, Radiat. Meas., 2020, 135, 106344.
28. X. Li, P. Zhai, X. Liu, G. Liu, R. Wei and H. Guo, Ceram. Int., 2024, 50, 20654–20663.
29. B. Bondzior, Opt. Lett., 2024, 49, 2125.
30. V. Nazabal, E. Fargin, B. Ferreira, G. Le Flem, B. Desbat, T. Buffeteau, M. Couzi, V. Rodriguez, S. Santran, L. Canioni and L. Sarger, J. Non-Cryst. Solids, 2001, 290, 73–85.
31. R. Reisfeld, L. Boehm and B. Barnett, J. Solid State Chem., 1975, 15, 140–150.
32. L. Skuja, J. Non-Cryst. Solids, 1992, 149, 77–95.
33. B. M. J. Smets, J. G. Verlijsdonk and J. Rutten, Mater. Res. Bull., 1989, 24, 431–439.
34. A. C. van der Steen and T. J. A. Aalders, Phys. Status Solidi B, 1981, 103, 803–808.
35. A. Scacco and P. W. M. Jacobs, J. Lumin., 1982, 26, 393–409.
36. P. W. M. Jacobs, Y. Kamishina, L. L. Coatsworth and M. J. Stillman, J. Lumin., 1979, 18–19, 619–622.
37. O. V. Laguta and I. M. Razdobreev, Sci. Rep., 2021, 11, 7766.
38. P. Rajanikanth, Y. Gandhi and N. Veeraiah, Opt. Mater., 2015, 48, 51–58.
39. C. H. Hsu, C. L. Liaw and C. H. Lu, J. Alloys Compd., 2010, 489, 445–450.
40. J. Huang, Y. Ni, Y. Ma, Y. Li, Z. Sun, X. Zhu, R. Wang, T. Li, X. Xi, G. Huang, T. Zhou, M. Li, H. Ren, L. Zhang and H. Chen, Ceram. Int., 2021, 47, 11415–11422.
41. M. L. Meretska, G. Vissenberg, A. Lagendijk, W. L. Ijzerman and W. L. Vos, ACS Photonics, 2019, 6, 3070–3075.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4dt03224a

---