JBCS

INTRODUCTION

The study of optical properties such as absorption, luminescence, and in particular, the energy transfer processes in lanthanide-ion-doped materials is fundamental. This is due to their potential applications in, for example, bio-sensing and bio-imaging,¹ photovoltaics,² and solid state lasers.³ The downconversion (DC) and upconversion (UC) processes have been widely studied, starting with the first theoretical proposals by Dexter⁴ and Auzel,⁵ respectively. The UC luminescence in Yb³⁺-Tb³⁺ co-doped glasses was first observed by Livanova et al. in 1969.⁶ This ion-pair interaction between Yb³⁺ (donor) and Tb³⁺ (acceptor) is known to lead to a simultaneous energy conversion of two near-infrared (NIR) photons into one visible photon after pumping the Yb³⁺:²F_5/2 level. The photon energy conversion has been explained in terms of a cooperative energy transfer (CET) process attributed to 2Yb³⁺:²F_5/2→²F_7/2; Tb³⁺:⁷F₆→⁵D₄.^7-9 This is also a non-resonant phonon-assisted process, confirmed by the study of the dependence of the Tb³⁺ UC visible luminescence with temperatures in the range of 10-300 K.^10,11 The Yb³⁺ ions are used as donors both due to the high absorption cross-section, and because the Yb³⁺:²F_5/2 level acts as a population reservoir when pumping with 980 nm.¹² These characteristics are important in regard to the UC process.

Among the various existing options, calcium lithium borate (CaLiBO) glasses are promising hosts for examining the effects of chemical environments on the optical properties of the lanthanides ions, due to their high transparency, low melting point, and high thermal stability.¹³ Another important characteristic of these glasses is their acceptance for a large concentration of lanthanides ions. Some authors have studied the energy transfer process in Dy³⁺-Ho³⁺, Tb³⁺-Er³⁺, Sm³⁺-Eu³⁺, Tb³⁺-Ho³⁺, Dy³⁺-Nd³⁺, Eu³⁺-Pr³⁺, and Eu³⁺-Ho³⁺ co-doped CaLiBO glasses.^13-19 We have studied the DC process in Tb³⁺-Yb³⁺ co-doped CaLiBO glasses.²⁰ However, to the best of our knowledge, the UC process in Yb³⁺-Tb³⁺ co-doped CaLiBO glasses has not yet been investigated. Moreover, a scarce spectroscopic analysis in Tb³⁺ doped CaLiBO glass is reported^21-23 and a few works reports on the determination of energy transfer probabilities from Yb³⁺ to Tb³⁺. The novelty of this work is the unequivocal determination of these values through the study of the UC luminescence kinetics. Therefore, CaLiBO glass was selected as the host material for the analysis of the UC processes in the Yb³⁺-Tb³⁺ system, prioritizing the studies of the influence of the variation of dopant concentrations on the system's optical properties.

A set of Yb³⁺-Tb³⁺ co-doped CaLiBO glasses with Yb³⁺ concentration from (0 to 7)% mol and with Tb³⁺ concentration fixed at 0.5% mol were studied. The spectroscopic characteristics of the glasses were investigated by absorption, luminescence and time-resolved spectroscopy measurements. The upconversion rise and decay times, energy transfer probability from Yb³⁺ to Tb³⁺ ions, and the efficiency of the processes that depopulate the Tb³⁺:⁵D₄ level after pumping the Yb³⁺:⁷F_5/2 level using a 5-ns pulsed laser as the excitation source were estimated.

EXPERIMENTAL

Synthesis of glasses

A CaLiBO glass matrix (30CaO + 10Li₂O + 60B₂O₃) (% mol) was prepared by the conventional melt-quenching method using high-purity CaCO₃, Li₂CO₃, and B₂O_3; this glass was then doped with Tb₄O₇ and Yb₂O₃, at approximately 1200 ºC for 1 h. The set of CaLiBO glasses doped with Tb₄O₇ and Yb₂O₃ were prepared with the following compositions: (99.5-x)% CaLiBO + 0.5% Tb₄O₇ + x Yb₂O₃, with x = 0, 0.1, 0.5, 1, 2.5, 5, and 7 (% mol). The glasses were cut and polished into a plate shape with the following dimensions: 3 mm × 10 mm × 10 mm.

Optical characterization

Ground state absorption spectra for both sets of samples were obtained using a Perkin Elmer Lambda 900 spectrophotometer in the 250-2500 nm spectral range. The visible Stokes luminescence spectra were obtained using a He-Cd laser (Kimmon/IK5652R-G) at 325 nm as the excitation source. The luminescent signal was dispersed by a monochromator (0.3 m, Thermo Jarrel Ash/82497), detected by a photomultiplier tube (PMT) (Hamamatsu/R928), and amplified by a lock-in amplifier. The UC luminescence spectra were obtained using an InGaAs index-guided strained-layer diode laser (Coherent, 980 nm excitation wavelength) as the excitation source, and the same laser was employed for the excitation power dependent measurements. Tb³⁺:⁵D₄→⁷F₅ UC luminescence kinetic response measurements were performed using an optical parametrical oscillator (OPO) (Surelite SLOP/Continumm) pumped by the third harmonic (355 nm) of a Nd-YAG laser (Surelite SLII-10/Continumm, 10 Hz, 5 ns) to generate excitation at a wavelength of 980 nm. The signal was dispersed by the same single monochromator (0.3 m) and the Tb³⁺:⁵D₄→⁷F₅ UC luminescence decay curves were recorded using the PMT and a digital oscilloscope (TekTronix/TDS380) to estimate the Tb³⁺:⁵D₄→⁷F₅ UC luminescence rise and decay times. All spectroscopic measurements were obtained at room temperature.

RESULTS AND DISCUSSION

Absorption spectra and Judd-Ofelt analysis

Figure 1 presents the absorption spectrum of 1.0Yb+2.0Tb codoped CaLiBO glass in the 250-1500 nm range. This spectrum is also representative for all the samples. The spectrum presents the typical Tb³⁺ bands in regards to line shapes and wavelengths. The electronic levels of Tb³⁺, excited by transitions from ⁷F₆ ground state are indicated, as well as the ²F_7/2→²F_5/2 transition of Yb³⁺ at 980 nm. The absorption coefficient in the NIR referring to the ²F_7/2→²F_5/2 transition of Yb³⁺ (980 nm) as a function of the nominal Yb³⁺ concentration had a linear dependence, and it is an indication of the successful incorporation of Yb³⁺ ions, for the prepared glass samples (not shown here).

Figure 1. Room temperature representative ground state absorption spectrum of 1.0Yb+2.0Tb co-doped glass sample. The transitions from ground states ⁷F₆ and ²F_7/2 are indicated

In order to analyze the radiative properties of the Tb³⁺ ions, such as the total radiative transition rate between ions, luminescence branching ratios and radiative lifetimes, the well-established Judd-Ofelt (JO)^24,25 theory was applied. The Ω_λ intensity parameters (λ = 2, 4, 6) are extracted by the experimental and calculated oscillator strengths for each ground state absorption transition.

The experimental oscillator strength is obtained from the absorption spectrum according to the following equation (1):

The oscillator strength calculated contains the reduce matrix element of the given transition, and the Ω_λ parameters, and is given by the equation (2):

In these equations, m is the electron mass, c is the speed of light, e is the electron charge, N is the number of doping ions per cm³, ∫α(ν) is the area of the absorption band, n is the refractive index, h is the Planck constant, ħ is the reduced Planck constant, ν is the transition frequency, and J is the ground state quantum number. The first term in parentheses of equation (2) accounts for the electric dipole contributions, where 〈aJ||U^λ||bJ'〉 are the reduced matrix elements of the tensor operator U^λ of rank λ, corresponding to transitions from 〈aJ| to |bJ'〉 to states. The matrix elements do not depend on the chemical environment and can be found in the literature.²⁶ The second term corresponds to the magnetic dipole contributions, with L + 2S being the magnetic dipole operator.

Once the intensity parameters are determined, they can be used to calculate the total radiative transition rate between two levels, a and b, which can be obtained through the following expression:

In which is the average transition energy in cm^-1 and J is the total angular momentum of the emitting level. The luminescence branching ratios and the radiative lifetime are given by:

and

respectively.

The absorption spectrum obtained for a sample with 2.0Tb was used to obtain the Ω_λ intensities parameters, where λ = 2, 4, 6 for the CaLiBO glass doped with Tb³⁺ ions.

Table 1 presents the experimental and calculated oscillator strengths for the J → J´ absorption transitions from the ground ⁷F₆ state to the excited levels, obtained using equations 1 and 2. The root-mean square (RMS) error for oscillator strengths was 1.8%. We determined the intensity parameters Ω_λ for CaLiBO:Tb³⁺: Ω₂ = 15.5 x 10^-20 cm², Ω₄ = 1.90 x 10^-20 cm² and Ω₆ = 3.69 x 10^-20 cm².

The radiative transition rates considering the electric and magnetic dipole contributions, branching ratios and radiative lifetimes of the ⁵D₄ and ⁵D₃ levels are given in Tables 2 and 3, respectively.

Luminescence

The visible Stokes luminescence spectra were obtained after direct excitation into the ⁵D₁ energy level of Tb³⁺ ions at 325 nm. Figure 2 shows the corresponding spectrum for the 1.0Yb+0.5Tb codoped glass. Regarding the line shape and wavelength positions, this spectrum is representative for the glasses with different Yb³⁺ doping concentrations. As observed, a dominant green emission band with a peak centered at approximately 544 nm was identified and assigned to the Tb³⁺:⁵D₄→⁷F₅ emitter level. The Tb³⁺:⁵D₄→⁷F_6,4,3 emitter levels were also identified with peak emission bands centered at 488, 586 and 622 nm, respectively.

Figure 2. Visible Stokes luminescence spectrum of the 1.0Yb+0.5Tb co-doped CaLiBO glass under λ_Exc=325 nm excitation

The visible Stokes luminescence is reduced by a factor of 2 with an increase in the concentration of Yb³⁺ ions from (0.1 to 2.5) %mol, and NIR luminescence in the 950-1100 nm range from Yb³⁺ ions is enhanced nearly 17 times (not shown here). This result is explained by the DC process from Tb³⁺ to Yb³⁺, as discussed in our previous article.²⁰ In that work, for the 0.0Yb+0.5Tb single-doped glass, after excitation at 488 nm and monitoring the Tb³⁺:⁵D₄ level at 544 nm, decay time was significantly higher, 2.38 ms than the Tb³⁺: ⁵D₄ level radiative lifetime, 1.87 ms obtained in this work using Judd-Ofelt calculations (see Table 2). This difference in times could be explained in terms of an energy migration between Tb³⁺ ions, which induces a lengthening on the ⁵D₄ emitter level.

Upconversion emission studies

An analysis of the UC luminescence for the Yb³⁺-Tb³⁺ co-doped CaLiBO glasses with excitation at 980 nm was performed. Figure 3 shows a simplified energy level diagram of the Yb³⁺ and Tb³⁺ ions with the relevant optical transitions. The UC through cooperative energy transfer (CET), the Yb³⁺-Tb³⁺ cross-relaxation (CR) and (ESA) excited state absorption processes are indicated by dotted lines. The non-radiative relaxation processes, excitation and emission transitions are indicated by dashed, dashed-dotted and solid lines, respectively.

Figure 3. Energy level diagrams of Yb³⁺ and Tb³⁺ ions. The UC through cooperative energy transfer (CET), Yb³⁺-Tb³⁺ cross-relaxation (CR), and excited state absorption (ESA) processes are indicated by dotted lines. The non-radiative relaxation processes, excitation and emission transitions are indicated by dashed, dashed-dotted and solid lines, respectively

The UC luminescence spectrum of the 1.0Yb+0.5Tb co-doped sample is presented in Figure 4 (a). The position and shape of the emission bands are consistent with those obtained upon direct excitation at 325 nm of the Tb³⁺:⁵D₁ level (Figure 3). A quadratic dependence on the excitation power at l_exc=980 nm was observed in a log-log plot (not shown here) for the UC emission bands associated with the Tb³⁺:⁵D₄→⁷F_J, J = 3, 4, 5, and 6 emitter levels. This indicates that the absorption of two NIR photons results in the emission of one visible photon, as shown in Figure 3. Is important to mention that the CET process is not resonant. This process is mediated by phonons (thermal energy) to bridge the electronic energy gap between the Tb³⁺: ⁵D₄ level and two times the lowest crystal field level of Yb³⁺: ²F_5/2.^10,11

Figure 4. (a) Tb³⁺ UC luminescence spectrum in the visible region of the 1.0Yb+0.5Tb co-doped CaLiBO glass upon excitation at λ_Exc=980 nm. The inset presents the dependence of the peak intensity with Yb³⁺ doping concentration of the emission band centered at 544 nm (Tb³⁺:⁵D₄→⁷F₅ emitter level) via the UC process. (b) UV-blue UC luminescence spectra for the samples xYb+0.5Tb with (x=0.5, 1.0, 7.0). The Tb³⁺:⁵D₃→⁷F_J, J = 6, 5, 4, and 3 emitter levels are indicated, excitation at λ_Exc=980 nm

The dependence of the peak intensity with Yb³⁺ concentration of the emission band centered at 544 nm (Tb³⁺:⁵D₄→⁷F₅ emitter level) via the UC process is shown in the inset of Figure 4 (a). In this inset, the Tb³⁺:⁵D₄→⁷F₅ UC luminescence increases nearly 68 times, with Yb³⁺ doping up to 2.5 %mol due to the increase in the number of Yb³⁺ activator ions; however, there is a quenching in the intensity for higher Yb³⁺ concentrations (> 2.5 %mol, indicating losses due to the Yb³⁺-Tb³⁺ CR processes (see Figure 3). This is even if the energy transfer probability from Yb³⁺ to Tb³⁺, extracted from time-resolved measurements, increases with Yb³⁺ concentration (as will be shown below). Figure 4 (b) shows the UC luminescence spectra of the xYb+0.5Tb co-doped samples in the UV-blue region (monitoring the ⁵D₃ level). The Tb³⁺:⁵D₃→⁷F_{3, 4, 5, 6} transitions are indicated, and an increase of the UC intensity can be observed. Moreover, the CR and ESA processes (see Figure 3) are confirmed by the curves of UC intensity versus pump power, monitoring the Tb³⁺:⁵D₃→⁷F₄ transition (not shown here). The emission around 440 nm presents a slope coefficient of 3, confirming the three-photon absorption mechanism that excites the Tb³⁺:⁵D₃ level.

UC luminescence kinetic response of Tb³⁺

Time-resolved measurements enable the extraction of further information regarding the energy transfer process involved. We attempted to estimate the energy transfer probability from Yb³⁺ to Tb³⁺, and the efficiency of the processes that depopulate the Tb³⁺:⁵D₄ level after resonantly pumping the Yb³⁺:⁷F_5/2 level. The analysis of the temporal evolution of the luminescence, specifically the rising feature after a short pulse, is a well-known fingerprint of an energy transfer process that does not require light to proceed, and therefore continues after the laser pulse.¹⁰ Thus, the Tb³⁺:⁵D₄→⁷F₅ (544 nm) UC luminescence kinetics was studied as a function of the Yb³⁺ doping concentration. These measurements enable a description of the Tb³⁺ UC processes.

Figure 5 shows the normalized Tb³⁺:⁵D₄→⁷F₅ (544 nm) UC luminescence kinetics after pumping the Yb³⁺:⁷F_5/2 level at 980 nm using a 5-ns pulsed laser as excitation source. The inset in Figure 5 shows the first millisecond following the laser pulse. It should be noted that the rise part of the Tb³⁺:⁵D₄→⁷F₅ (544 nm) UC luminescence transient rose slowly from zero intensity after the laser pulse. The rise part of the transient was strongly affected by the Yb³⁺ concentration. In addition, the decay part of the transient exhibited non-exponential behavior.

Figure 5. Kinetic response of Tb³⁺ UC luminescence for the samples xYb+0.5Tb with (x = 0.1, 0.5, 1.0, 5.0, 7.0) upon excitation at λ_Exc = 980 nm, monitored at λ_Obs = 544 nm (Tb³⁺:⁵D₄→⁷F₅ emitter level). The inset shows the responses at the first millisecond after the laser pulse

The results indicate that the origin of the Tb³⁺:⁵D₄→⁷F₅ (544 nm) UC luminescence is due to a CET process from two Yb³⁺ ions to one Tb³⁺ ion. Thus, in accordance with Salley et al.,²⁷ the transients can be fitted using the following equation (6):

where τ₁ and τ₂ represent the decay and rise times of the transient, respectively. τ₁ provides information about the radiative and non-radiative process in the acceptor ion and the inverse of τ₂ (W_ET = 1/τ₂) provides the energy transfer probability from the donor to the acceptor ion. Note that we have successfully employed equation (6) to estimate the energy transfer probability between Tb³⁺ and Yb³⁺ ions in our previous work.¹¹

Therefore, τ₁ decreases from 1.64 to 0.79 ms and W_ET increases from 6.7 × 10³ s^-1 to 45.5 × 10³ s^-1 with Yb³⁺ doping in the range of 0.1-7.0 %mol (see Table 4). In the work reported by Terra et al.,¹¹ the mechanisms involved in the UC process in Yb³⁺+Tb³⁺ co-codoped aluminosilicate glasses were analyzed. In that work, the values reported for the UC luminescence rise times are in the range from 20 to 65 microseconds for the temperature range from 123 to 473 K. The inverse of those rise times results in values of probability of energy transfer in the range from 50 × 10³ to 15.4 × 10³ s^-1, which are comparable to the values calculated in the present work. On the other hand, in the work reported by Salley et al.,²⁷ studying the UC process in a Yb³⁺+ Tb³⁺ co-codoped SrCl₂ crystal, they obtained an energy transfer probability value of the order of 6.9 × 10³ s^-1, measured at 100 K. It can be seen that the W_ET values of this work and those of Terra are greater than that reported by Salley and this is understandable, since it is known that rare earth ions agglomerates are usually formed in glasses and not in crystals, which facilitates the energy transfer from one ion to another.

The non-exponential behavior of the decay part of the transient and the decrease of τ₁ with increasing Yb³⁺ concentration in this set of samples is due to the presence of non-radiative processes involved in the Tb³⁺:⁵D₄ level. The most likely non-radiative process that depopulates the Tb³⁺:⁵D₄ level after pumping the Yb³⁺:⁷F_5/2 level is the Yb³⁺-Tb³⁺ CR (as depicted in Figure 3).

For comparison purposes, the Tb³⁺:⁵D₄→⁷F₅ (544 nm) decay times after excitation at 488 nm (extracted from our previous DC experiment)²⁰ are shown in Table 4. Note that these decay times are considerably higher than those estimated in the UC experiment analyzed here. These results indicate that the CR process is present after excitation at 980 nm.

The efficiency of the CR process (Figure 3) that depopulates the Tb³⁺:⁵D₄ level after pumping the Yb³⁺:⁷F_5/2 level could be calculated using the following equation (7):

where τ_xYb+Tb is the Tb³⁺:⁵D₄→⁷F₅ (544 nm) UC luminescence decay time in the presence of Yb³⁺, and τ_Tb = 2.38 ms is the Tb³⁺:⁵D₄→⁷F₅ (544 nm) luminescence decay time for the Tb³⁺ singly doped sample (0.5 %mol) after excitation at 488 nm (see Ref²⁰). These values are summarized in Table 4, where η is observed to increase from 31% to 67% when the concentration of Yb³⁺ ions increases, indicating that the CR process is efficient.

We have observed that the Tb³⁺ UC luminescence slowly rises over time in our samples. The origin of the Tb³⁺ UC luminescence is due to a CET process in which two Yb³⁺ ions simultaneously transfer the energy to one Tb³⁺ ion (2Yb³⁺:²F_5/2; Tb³⁺:⁷F₆→2Yb³⁺:²F_7/2; Tb³⁺:⁵D₄), as depicted in Figure 3.

CONCLUSIONS

This paper presents a qualitative and quantitative perspective on the study of the spectroscopic properties of Yb³⁺-Tb³⁺ co-doped CaLiBO glasses, using absorption, luminescence and time-resolved measurements. Intensity parameters Ω_λ for CaLiBO:Tb³⁺ glass have been determined by the Judd-Ofelt method to be Ω₂ = 15,5 × 10^-20 cm², Ω₄ = 1,90 × 10^-20 cm² and Ω₆ = 3,69 × 10^-20 cm². These values were used to calculate spectroscopic properties, such as oscillator strengths, probabilities, and radiative lifetimes. The analyses covered herein showed that CaLiBO glass is a good host for high concentrations of Yb³⁺ and Tb³⁺ ions. The rise and decay times of the Tb³⁺:⁵D₄→⁷F₅ UC transient luminescence allowed for determination of the energy transfer probability and the efficiency of the processes that depopulate the Tb³⁺:⁵D₄ level after resonantly pumping the Yb³⁺:⁷F_5/2 level. The time-resolved measurements identified the origin of the Tb³⁺:⁵D₄→⁷F₅ UC luminescence as a cooperative energy transfer mechanism from two Yb³⁺ ions to one Tb³⁺ ion. The results suggest that Yb³⁺-Tb³⁺ codoped CaLiBO glasses are potential materials for applications in the form of optical converters.

ACKNOWLEDGMENTS

The authors acknowledge financial support from the following Brazilian Agencies: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq).

REFERENCES

1. Escudero, A.; Becerro, A. I.; Carrillo-Carrión, C.; Núñez, N. O.; Zyuzin, M. V.; Laguna, M.; González-Mancebo, D.; Ocaña, M.; Parak, W. J.; Nanophotonics 2017, 6, 1568.

2. Yang, W.; Li, X.; Chi, D.; Zhang, H.; Liu, X.; Nanotechnology 2014, 25, 482001.

3. Borrero-González, L. J.; Terra, I. A. A.; Nunes, L. A. O.; Farias, A. M.; Barboza, M. J.; Rohling, J. H.; Medina, A. N.; Baesso, M. L.; Appl. Phys. B: Lasers Opt. 2012, 107, 415.

4. Dexter, D. L.; Phys. Rev. 1957, 108, 630.

5. Auzel, F.; C. R. Seances Acad. Sci., Ser. B 1966, 263, 819.

6. Livanova, L. D.; Saitkulo, I. G.; Stolov, A. L.; Soviet Physics - Solid State 1969, 11, 750.

7. Ostermayer, F. W.; Van Uitert, L. G.; Phys. Rev. B 1970, 1, 4208.

8. Martín, I. R.; Yanes, A. C.; Méndez-Ramos, J.; Torres, M. E.; Rodríguez, V. D.; J. Appl. Phys. 2001, 89, 2520.

9. Huang, L.; Yamashita, T.; Jose, R.; Arai, Y.; Suzuki, T.; Ohishi, Y.; Appl. Phys. Lett. 2007, 90, 131116.

10. Salley, G. M.; Valiente, R.; Güdel, H. U.; J. Phys.: Condens. Matter. 2002, 14, 5461.

11. Terra, I. A. A.; Borrero-González, L. J.; Nunes, L. A. O.; Belançon, M. P.; Rohling, J. H.; Baesso, M. L.; Malta, O. L.; J. Appl. Phys. 2011, 110, 083108.

12. Terra, I. A. A.; de Camargo, A. S. S.; Terrile, M. C.; Nunes, L. A. O.; J. Lumin. 2008, 128, 891.

13. Tripathi, G.; Rai, V. K.; Rai, S. B.; Appl. Phys. B: Lasers Opt. 2006, 84, 459.

14. Tripathi, H. B.; Kandpal, H. C.; Agarwal, A. K.; Phys. Status Solidi A 1979, 52, 697.

15. Joshi, J. C.; Joshi, B. C.; Pandey, N. C.; Belwal, R.; Joshi, J.; J. Solid State Chem. 1977, 22, 439.

16. Joshi, J. C.; Pandey, N. C.; Joshi, B. C.; BelWal, R.; Joshi, J.; J. Non-Cryst. Solids 1978, 27, 173.

17. Joshi, J. C.; Joshi, B. C.; Pandey, N. C.; Pandey, B. C.; Joshi, J.; J. Solid State Chem. 1978, 26, 179.

18. Tripathi, H. B.; Agarwal, A. K.; Kandpal, H. C.; Belwal, R.; Solid State Commun. 1978, 28, 807.

19. Tripathi, H. B.; Kandpal, H. C.; Agarwal, A. K.; Belwal, R.; Chem. Phys. Lett. 1978, 57, 50.

20. Terra, I. A. A.; Borrero-González, L. J.; Figueredo, T. R.; Almeida, J. M. P.; Hernandes, A. C.; Nunes, L. A. O.; Malta, O. L.; J. Lumin. 2012, 132, 1678.

21. Yamashita, T.; Ohishi, Y.; J. Non-Cryst. Solids 2008, 354, 1883.

22. Hoshina, T.; Jpn. J. Appl. Phys. (1962-1981) 1967, 6, 1203.

23. Colak, S.; Zwicker, W. K.; J. Appl. Phys. 1983, 54, 2156.

24. Ofelt, G. S.; J. Chem. Phys. 1962, 37, 511.

25. Judd, B. R.; Phys. Rev. 1962, 127, 750.

26. Carnall, W. T.; Fields, P. R.; Rajnak, K.; J. Chem. Phys. 1968, 49, 4447.

27. Salley, G. M.; Valiente, R.; Guedel, H. U.; J. Lumin. 2001, 94, 305.