Synthesis, photoluminescence and Judd–Ofelt analysis of red LiGd₅P₂O₁₃ : Eu³⁺ phosphors for white LEDs

Abstract

A series of LiGd₅P₂O₁₃ : Eu³⁺ phosphors with a red-emitting band centered at 622 nm were synthesized by a high temperature solid-state reaction method. Structure-luminescence correlation, concentration quenching and thermal stability of the LiGd₅P₂O₁₃ : Eu³⁺ phosphors were studied combined with XRD, PL/PLE spectra as well as temperature-dependent PL and decay curves. The optimum concentration of Eu³⁺ doped in the LiGd₅P₂O₁₃ is 20 mol%, and the corresponding concentration quenching mechanism is found to be via dipole/quadrupole–quadrupole interactions. Intensity parameters (Ω₂ and Ω₄) and various radiative properties such as transition rates (A), branching ratios (β) and stimulated emission cross-section (δ_e) were calculated using the Judd–Ofelt theory. LiGd₅P₂O₁₃ : Eu³⁺ exhibited good thermal stability and its emission intensity decreased slightly at temperature above 150 °C. The CIE coordinates of LiGd₅P₂O₁₃ : 0.20Eu³⁺ (0.644, 0.339) are very close to s-RGB standard red (0.640, 0.330), and the corresponding quantum efficiency is 46.5%. The results show that LGPO : Eu³⁺ may be considered as a potential red emitting phosphor for NUV/blue pumped WLEDs.

---

1. Introduction

Recently, white light-emitting diodes (WLEDs), which are regarded as a next-generation solid-state lighting source, have attracted more and more attention due to their outstanding merits in many aspects such as long operational lifetime, high brightness, low energy consumption and environmental friendliness.¹ Currently, the commercially available WLEDs are based on a blue chip with yellow phosphor (YAG : Ce). However, this combination does not contain a red component, which leads to a low color rendering index.² One alternative approach to obtain white light is using a near-UV (NUV) InGaN-based LED chip coated with blue/green/red tricolor phosphors.^3,4 However, the efficiency and thermal stability of the currently used red phosphor is relatively low compared to the other two components. Thus, there is an urgent need to develop efficient and stable red phosphors that have an excitation wavelength matching the emission wavelength of the NUV LEDs (λ_em = ∼395 nm) and blue LEDs (λ_em = ∼460 nm).

Eu³⁺ is an important rare earth ion in synthesizing phosphors with excellent properties, for example Y₂O₂S : Eu³⁺ is used as a red phosphor in fluorescent lamp. Recently, many novel Eu³⁺-doped compounds have been reported as promising candidates as potential red-emitting phosphors for LED application.^5,6 It is well-known that phosphates possess many excellent properties as phosphor hosts, i.e. low sintering temperature, large band gap and moderate phonon energy, high thermal and chemical stability.⁷ Some Eu³⁺ doped phosphate phosphors are reported to have fairly good luminescence properties, such as Ba₁₀F₂(PO₄)₆ : Eu³⁺,⁸ Ca₉R(PO₄)₇ : Eu³⁺ (R = Al, Lu)⁹ and Ca₁₉Mg₂(PO₄)₁₄ : Eu³⁺.¹⁰

The host LiGd₅P₂O₁₃ (abbreviated as LGPO) was firstly discovered by Chen's group in 2007.¹¹ LiGd₅P₂O₁₃ presents a new structural type and is built up from [Gd₅P₂O₁₃]⁻ layers and one-dimensional Li chains with an unusual Li–Li distance. However, up to now, there is no report on the red phosphors based on LGPO host by doping Eu³⁺ ions.

In this paper, a systematic investigation on photo-luminescence spectra, concentration quenching, as well as thermal quenching of LGPO : Eu³⁺ phosphor were carried out. Judd–Ofelt theory was applied to calculate the intensity parameters (Ω₂ and Ω₄) and various other radiative properties such as radiative transition rates, branching and asymmetry ratios, stimulated emission cross-section, which might provide a significant insight into the structure–luminescence correlation in LGPO : Eu³⁺ system. It is found that LGPO : Eu³⁺ is an efficient red phosphor with a high red color purity and good thermal stability. The results show that LGPO : Eu³⁺ may be considered as potential red emitting phosphor for NUV/blue pumped WLEDs.

2. Experimental

Li₂CO₃ (AR), Gd₂O₃ (AR), NH₄H₂PO₄ (AR) and Eu₂O₃ (4N) were used as starting materials to synthesize LiGd_5(1−x)P₂O₁₃ : xEu³⁺ (0.1 ≤ x ≤ 1.0) phosphors by solid state reaction. The initial materials were weighed according to the stoichiometric ratio, and mixed in an agate mortar in ethanol media with intermittent grinding and drying. The homogeneous mixture was preheated at 300 °C for 4 h, then calcined at 1150 °C for 12 h with a heating rate of 300 °C h⁻¹.

Crystalline phases of samples were identified by an X-ray diffraction (XRD, Model Rigaku-D/Max-2200PC, Japan) using Cu Kα radiation with a step size of 0.02° and a speed of 10° min⁻¹ within the range from 10° to 75°. The photoluminescence excitation (PLE)/emission (PL) spectra and luminescence decay curves of samples were measured on a Edinburgh FLS-920 Time Resolved and Steady State Fluorescence Spectrometer with a 450 W xenon lamp as an excitation light source. The internal quantum efficiency was also obtained from FLS-920 spectrometer with integrated sphere. The CIE coordinates are calculated by a software, which was developed by M.H. Chan, Department of Physics, Hong Kong Baptist University (HKBU). All the measurements were recorded at room temperature.

3. Results and discussions

3.1 XRD and structure

Fig. 1a illustrates the X-ray diffraction (XRD) patterns of LGPO : xEu³⁺ (x = 0.2, 0.4, 0.6 and 0.8) phosphors with different doping Eu³⁺ contents (x). All the observed diffraction peaks can be indexed to the standard data of LiGd₅P₂O₁₃ reported by Chen's group.¹¹ No obvious impurity phase was detected when Eu³⁺ ions were doped into the host lattice, indicating that all samples are of single phase.

Fig. 1 XRD patterns of LGPO : xEu³⁺ (x = 0.2, 0.4, 0.6 and 0.8) phosphors (a), and the crystal structure of LiGd₅P₂O₁₃ with Gd polyhedra (b).

The structure of LiGd₅P₂O₁₃ belongs to the centrosymmetric space group C2/m with unit cell parameters a = 18.645 Å, b = 5.625 Å, c = 12.014 Å, β = 117.55°, V = 1117.3 ų, and Z = 4. As shown in Fig. 1b, the three-dimensional (3D) framework of LiGd₅P₂O₁₃ consists of isolated PO₄ tetrahedra linked with Gd atoms distributed among them, and Li atoms are located in the infinite tunnels along the b-axis which are delimited by PO₄ tetrahedra and Gd polyhedra. There are five Gd polyhedra in LGPO host, i.e. Gd(1) and Gd(5) are six-coordinated, and Gd(2), Gd(3), Gd(4) are seven-coordinate. All Gd–O distances vary in the range from 2.206 to 2.605 Å, and all Gd sites possess the point symmetry of C₁. Due to the similarity of charge and ionic radius (Gd³⁺ (r = 0.94 Å) and Eu³⁺ (r = 0.95 Å)¹²), it is reasonable to purpose the substitution of Gd³⁺ by Eu³⁺. Since Eu³⁺ ions occupy highly asymmetric sites, a red emission (⁵D₀–⁷F₂) whose intensity is much stronger than that of orange emission (⁵D₀–⁷F₁) is expected in LGPO : xEu³⁺ phosphors.

3.2 PLE and PL spectra

The PLE and PL spectra of LGPO : 0.2Eu³⁺ are shown in Fig. 2a. The PLE spectrum contains a series of narrow bands locating in the range of 240–480 nm. These narrow bands could be attributed to the typical 4f–4f transition absorption bands of Eu³⁺, in which the two intense bands centered at 395 nm and 467 nm are attributed to ⁷F₀–⁵L₆ and ⁷F₀–⁵D₂ transition. This demonstrates that the phosphor can be excited efficiently by near-ultraviolet and blue light. The other excitation peaks centered at 319 nm, 362 nm, 376 nm, 383 nm, 398 nm, and 415 nm correspond to the transitions from ⁷F₀ to ⁵H₆, ⁵D₄, ⁵G₃, ⁵L₇, ⁵L₆, and ⁵D₃ levels, respectively.

Fig. 2 The PLE and PL spectra of LGPO : 0.2Eu³⁺ (a), and corresponding CIE chromaticity diagram with image under a 365 nm UV lamp (b).

As shown in Fig. 2a, the PL spectra are composed of four bands centered at 579, 593, 622, 652, and 706 nm assigned to the ⁵D₀–⁷F_J (J = 0, 1, 2, 3, and 4) transitions of Eu³⁺, respectively. It is known that ⁵D₀–⁷F₀ transition is forbidden and sensitive to the crystal field, which exists only when Eu³⁺ occupies sites with local symmetries of C_n, C_nv or C_s.¹³ The ⁵D₀–⁷F₁ transition peak originates from magnetic dipole transition. In contrast, the radiative transitions from ⁵D₀ to ⁷F₂ and ⁷F₄ levels were electric dipole in character. The ⁵D₀–⁷F₃ transition was forbidden from both electric and magnetic dipole considerations.¹⁴ The ⁵D₀–⁷F_J emission of the Eu³⁺ ion has a close relationship with its occupied chemical environment. When the Eu³⁺ ion was located at site with inversion symmetry, the magnetic dipole (⁵D₀–⁷F₁) emission transition was dominant, whereas at site without inversion symmetry, the electric dipole (⁵D₀–⁷F₂) transition becomes dominant.¹⁵ Thus, the intensity ratio of ⁵D₀–⁷F₂ to ⁵D₀–⁷F₁, R = I₂/I₁, is a good way to detect the symmetry of the coordinated environment around Eu³⁺ ion. A low symmetry leads to a high value of R (>1), the opposite will lead to a low value (1 > R > 0). The intensity of I₂ and I₁ are defined as the area under their corresponding emission spectrum curves calculated by integrating from 606 to 640 nm and 585 to 600 nm, respectively. The asymmetry ratio R of representative sample LGPO : 0.2Eu³⁺ is 7.236, which indicates that Eu³⁺ occupies the sites with no inversion symmetry, and agrees well with the result of the above structural discussion.

The CIE coordinates of LiGd₅P₂O₁₃ : 0.2Eu³⁺ phosphors are calculated to be x = 0.644, y = 0.339, as shown in Fig. 2b. The CIE coordinate of LiGd₅P₂O₁₃ : 0.2Eu³⁺ is very close to the CIE coordinate of s-RGB standard red (x = 0.64, y = 0.33), and is considered to be better than that of the commercial Y₂O₂S : Eu³⁺ (x = 0.622, y = 0.351).

3.3 Concentration quenching

Fig. 3 shows the concentration dependence of PL spectra of LGPO : xEu³⁺ samples (x = 0.1–1.0) under 395 nm excitation. LGPO : xEu³⁺ samples have the same spectral profile with the different Eu³⁺ concentration. However, the luminescence intensity increases with the doping concentrations up to 0.2. With further increase of Eu³⁺ concentration, the emission intensity decreases. The distance between the Eu³⁺ ions becomes small as the concentrations of Eu³⁺ increases; thus the probability of energy migration increases. The concentration quenching phenomena will not occur if the average distance between identical Eu³⁺ ions is so large that the energy migration is hampered.

Fig. 3 Concentration dependence of PL spectra of LGPO : xEu³⁺ samples (λ_ex = 467 nm).

Generally speaking, the energy transfer process usually originates from exchange interaction, radiation re-absorption or multipole–multipole interaction. To investigate the concentration quenching phenomena of the phosphor, the critical distance (R_c) between Eu³⁺ ions for energy transfer can be calculated using the relation proposed by Blasse and Grabmaier as follows:¹⁶

where V is the volume of the unit cell, x_c is the critical concentration of the activator ion, and N is the number of formula unit per unit cell. In the case of LiGd_5(1−x)P₂O₁₃ : xEu³⁺ phosphors, x_c = 0.20, V = 1117.30 ų, and N = 4. Therefore, R_c was calculated to be 13.86 Å. Since the corresponding critical distance for exchange interaction is about 5–8 Å, and there is tiny overlap between PL and PLE spectrum in present case, it is deduced that the electric multipolar interaction is involved in energy transfer.

To further validate the concentration quenching effect of Eu³⁺ in LiGd₅P₂O₁₃ : Eu³⁺, the decay lifetimes of Eu³⁺ were investigated. When there is no interaction between rare earth ions, the decay curve is usually a single exponential function. But the decay curves of LGPO : xEu³⁺ (x > 0.2) samples exhibit obvious deviations from the single exponential decay. Thus, the average decay time (τ) was determined by the following formula:

As shown in Fig. 4a and Table 1, Decay time of ⁵D₀–⁷F₂ transition decreases as Eu³⁺ contents increase. The ⁵D₀–⁷F₂ decay time of LGPO : 0.1Eu³⁺ is 1260.46 μs, whereas that of LGPO : 1.0Eu³⁺ was 62.60 μs. The similar tendency happens in the decay time of ⁵D₀–⁷F₄ transition (Fig. 4b), which is a strong proof of enhanced non-radiative energy transfer between Eu³⁺ ions with shorten Eu³⁺–Eu³⁺ distance. The decay time of the optimized-composition LGPO : 0.2Eu³⁺ phosphor seems to be quite long (>1 ms), which could cause phosphor quenching from high LED flux if encapsulating phosphors directly on LED chip. For the phosphors with long decay time, it is suggested that they are best used in remote phosphor package, in which a phosphor composite is precisely layered onto a substrate and separated from the LED energy source, in order to reduce the incident flux on the phosphor.¹⁷

Fig. 4 Decay curves of the Eu³⁺ fluorescence in LGPO : xEu³⁺ samples (excited at 467 nm, monitored at 622/706 nm).

Table 1 The average decay lifetimes of ⁵D₀–⁷F_2,4 transition in LGPO : xEu³⁺ samples

Compounds	^5D0–^7F2 decay lifetime (μs)	^5D0–^7F4 decay lifetime (μs)
LGPO : 0.1Eu^3+	1260.46	1210.08
LGPO : 0.2Eu^3+	885.53	879.57
LGPO : 0.3Eu^3+	579.87	574.05
LGPO : 0.4Eu^3+	379.57	351.85
LGPO : 0.5Eu^3+	260.23	249.02
LGPO : 0.6Eu^3+	105.23	101.40
LGPO : 0.7Eu^3+	100.74	95.96
LGPO : 0.8Eu^3+	62.60	58.63
LGPO : 1.0Eu^3+	46.72	42.94

---

There are three type of electric multipolar interaction, i.e. dipole–dipole (d–d), dipole–quadrupole (d–q) and quadrupole–quadrupole (q–q) interactions. Thus, there is a need to elucidate which type of interaction is involved in energy transfer. According to Dexter's theory, the relation between luminescent intensity (I) and activator concentration (x) can be expressed by this equation:¹⁸

where x is the activator concentration, I/x is the emission intensity (I) per activator concentration (x), and K and β are constants for the same excitation condition for a given host crystal. Q = 3 stands for energy transfer among the nearest neighbor ions, while Q = 6, 8 and 10 stands for dipole–dipole, dipole–quadrupole, and quadrupole–quadrupole interactions.¹⁹Eqn (3) can thus be represented by the follow equation:

As illustrated in Fig. 5, the relationship of log(τ/x) versus log(x), using either ⁵D₀–⁷F₂ or ⁵D₀–⁷F₄ decay times, are linear. And the slope of the line is about −3. So the value of Q is approximately equal to 9, which clearly indicates that both dipole–quadrupole and quadrupole–quadrupole interactions play part in the concentration quenching process in LGPO : Eu³⁺.

Fig. 5 The dependence of log(τ/x) on log(x) of LGPO : xEu³⁺ samples.

3.4 Judd–Ofelt parameters and radiative properties

The Judd–Ofelt (J–O) theory has been successfully applied to the quantitative determination of the optical properties of trivalent lanthanides.²⁰ The Judd–Ofelt intensity parameters (Ω_2,4) provide valuable information about the local structure and bonding in the vicinity of rare earth ions. By analyzing the data from emission spectra and decay lifetimes, the local environment around the metal ion and the bond covalency of metal–ligand bonds can be interpreted.²¹

As a representative sample, the data of LiGd₅P₂O₁₃ : 0.2Eu³⁺ was analyzed by Judd–Ofelt theory. The total relative integrated intensity I is proportional to the total radiative transition rate which is equal to the reciprocal of experimental lifetime τ. Therefore, the equation could be written as:

where a is a constant. The branching ratio β corresponding to the emission from an excited level to its lower levels is as follows:

Thus, each ⁵D₀–⁷F_J (J = 0, 1, 2, 3 and 4) radiative transition rate was calculated and shown in Table 2.

Table 2 Judd–Ofelt parameters (transition rates, branching ratios, Ω₂ and Ω₄), effective bandwidth of the emission transition (Δλ_eff), stimulated emission cross-section (δ_e) in LiGd₅P₂O₁₃ : 0.2Eu³⁺

LiGd5P2O13 : 0.2Eu^3+
Transition	Wavenumber (cm^−1)	Transition rate (s^−1)	β (%)	Ω J (× 10^−20 cm^2)	Δλeff (nm)	δ e (× 10^−22 cm^2)
^5D0–^7F0	17256.25	31.964	2.83		3.25	3.640
^5D0–^7F1	16849.19	95.496	8.45		8.00	4.861
^5D0–^7F2	16090.10	691.00	61.15	8.793	18.5	18.29
^5D0–^7F3	15313.93	35.566	3.15		7.95	2.652
^5D0–^7F4	14164.30	251.84	22.29	6.536	7.25	28.33

---

The radiative transition rates of electric dipole transitions could be expressed as a function of the J–O intensity parameters:²²

where n is the refractive index of host, h is Planck's constant (h = 6.626 × 10⁻²⁷ erg s), ν is the wave-number of radiative transition, and e is the charge of electron (e = 4.803 × 10⁻¹⁰ esu). The values are the square reduced matrix elements which are independent of the chemical environment of the Eu³⁺ ion.

The refractive index (n) could be calculated by following equation:²³

where S_md is the magnetic dipole line strength (S_md = 7.83 × 10⁻⁴²). Therefore, a value of 2.01 for refractive index (n) could be obtained.

For LiGd₅P₂O₁₃ : 0.2Eu³⁺, the value of Ω₂ and Ω₄ is found as 8.793 × 10⁻²⁰ and 6.536 × 10⁻²⁰ cm², respectively. It is known that Ω₂ is related to the covalency and structural changes in the vicinity of the Eu³⁺ ion (short range effect) while Ω₄ is not directly related to the symmetry of the Eu³⁺ ion but to the electron density on the surrounding ligands (long range effect).²⁴ The great individual values of Ω₂ and Ω₄ indicate the great extent of the covalent character/electron density around the Eu³⁺ ion, and the high value of the ratio Ω₂/Ω₄ (>1) indicates the presence of Eu³⁺ in a highly asymmetric environment. The calculation results support the above-mentioned structural discussion and experimental observation of optical properties.

The stimulated emission cross-section (δ_e) was an essential parameter to signifies the rate of energy extraction from the lasing material and is calculated as:²³

where λ_p is the emission peak wavelength, c is the velocity of light and Δλ_eff is the effective bandwidth of the emission transition. Effective bandwidth of the emission transition (Δλ_eff) and stimulated emission cross-section (δ_e) for each ⁵D₀–⁷F_J transitions are shown in Table 2. The product of emission cross-section and the effective bandwidth of the emission transition is a significant parameter to predict the bandwidth of the optical amplifier. The higher the product values were, the better was the amplifiers performance. LiGd₅P₂O₁₃ : 0.2Eu³⁺ have comparable δ_e values of ⁵D₀–⁷F_J transitions with those of reported BiOF : Eu³⁺, indicating that it has some potential for red laser applications.

3.5 Temperature-dependent PL properties and quantum efficiency

The phosphor layers in LED devices are expected to operate at elevated temperatures (∼150 °C).²⁵ Therefore, the thermal quenching of LED phosphors is one of the most important technological parameters because it has great influence on the light output and CRI of devices. Thus, temperature-dependent PL spectra under 467 nm excitation of as-prepared LiGd₅P₂O₁₃ : 0.2Eu³⁺ are indicated in Fig. 6. The relative peak intensity (excitation at 467 nm and monitored at 622 nm) of LGPO : 0.2Eu³⁺ decreased marginally with rising temperature. It reached 80% of initial value at 100 °C and then 70% at 150 °C, which is better than reported Sr_1.7Zn_0.3CeO₄ : Eu³⁺ (20% at 150 °C)²⁶ and Ba₂ZnB₂O₆ : Eu³⁺ (30% at 150 °C).²⁷ The fair thermal stability demonstrated that LiGd₅P₂O₁₃ : Eu³⁺ red phosphor could be potential for high-powered LED applications. The internal quantum efficiency value (IQE) of LGPO : 0.20Eu³⁺ under 395 and 467 excitation is 46.5%, which is comparable with previous reported Sr_1.7Zn_0.3CeO₄ : Eu³⁺ (46.1%),²⁶ and lower than commercialized red nitride phosphor Sr₂Si₅N₈ : Eu²⁺ (78.2%).²⁶ However, the QE of LGPO : 0.20Eu³⁺ phosphor may be further boosted by optimizing the synthesis process.

Fig. 6 Temperature-dependent PL spectra of LGPO : 0.2Eu³⁺ phosphor (λ_ex = 467 nm).

4. Conclusions

In summary, the LiGd₅P₂O₁₃ : Eu³⁺ phosphors were synthesized and systematically investigated. The phosphor could be excited efficiently by near-ultraviolet and blue light, and emits strong and high-purity red light. The CIE coordinate is calculated to be (0.644, 0.339). The optimum concentration of Eu³⁺ for LiGd_5(1−x)P₂O₁₃ : xEu³⁺ is determined to be x = 0.2. By analyzing PL spectra and decay lifetimes, the critical distance (R_c) and the concentration quenching mechanism is found to be 13.86 Å and dipole–quadrupole–quadrupole interaction, respectively. From the Judd–Ofelt analysis, it is observed that the intensity parameter Ω₂ was greater than Ω₄. Which suggest high covalency from the metal (Eu) to ligands (O), and the high value of the ratio Ω₂/Ω₄ (>1) indicates the presence of Eu³⁺ in a highly asymmetric environment. LiGd₅P₂O₁₃ : Eu³⁺ exhibited efficient red emission (IQE = ∼50%) and good thermal stability (70% at 150 °C). The results show that LGPO : Eu³⁺ may be considered as potential red emitting phosphor for NUV/blue pumped WLEDs.

Acknowledgements

This research was supported by Guangxi Natural Science Foundation (Grant no. 2014GXNSFBA118046) and National Natural Science Foundation of China (no. 61264003).

References

1. Q. Zhou, Y. Zhou, Y. Liu, L. Luo, Z. Wang, J. Peng, J. Yan and M. Wu, J. Mater. Chem. C, 2015, 3, 3055.
2. J. Zhou, F. Huang, J. Xu, H. Chen and Y. Wang, J. Mater. Chem. C, 2015, 3, 3023.
3. N. Komuro, M. Mikami, Y. Shimomura, E. G. Bithell and A. K. Cheetham, J. Mater. Chem. C, 2015, 3, 204.
4. Q. Liu, Z. Zheng, X. Zhang and Z. Bai, J. Alloys Compd., 2015, 628, 298.
5. H. Zhou, Y. Jin, M. Jiang, Q. Wang and X. Jiang, Dalton Trans., 2015, 44, 1102.
6. X. Zhang, J. Zhang and M. Gong, Opt. Mater., 2014, 36, 850.
7. X. Zhang, C. Zhou, J. Song, L. Zhou and M. Gong, J. Alloys Compd., 2014, 592, 283.
8. Y. Peng, W. Shi, C. Han, Y. Kang, Y. Wang and Z. Zhang, Spectrochim. Acta, Part A, 2015, 145, 194.
9. F. Du, Y. Nakai, T. Tsuboi, Y. Huang and H. Seo, J. Mater. Chem., 2011, 21, 4669.
10. G. Zhu, Z. Ci, Y. Shi, M. Que, Q. Wang and Y. Wang, J. Mater. Chem. C, 2013, 1, 5960.
11. J. Zhu, W. Cheng, D. Wu, H. Zhang, Y. Gong, H. Tong and D. Zhao, Inorg. Chem., 2007, 46, 208.
12. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
13. X. Bai, G. Zhang and P. Fu, J. Solid State Chem., 2007, 180, 1792.
14. X. Dong, J. Zhang, X. Zhang, Z. Hao and Y. Luo, J. Alloys Compd., 2014, 587, 493.
15. N. Zhang, C. Guo and H. Jing, RSC Adv., 2013, 3, 7495.
16. X. Zhang, F. Mo, L. Zhou and M. Gong, J. Alloys Compd., 2013, 575, 314.
17. Y. Zhu and N. Narendran, J. Light Visual Environ., 2008, 32, 115–119.
18. D. Dexter and J. Schulman, J. Chem. Phys., 1954, 22, 1063.
19. C. Huang, L. Luo and T. Chen, J. Electrochem. Soc., 2011, 158, J341.
20. M. H. V. Werts, R. T. F. Jukes and J. W. Verhoeven, Phys. Chem. Chem. Phys., 2002, 4, 1542.
21. C. R. Raju, C. A. Reddy, S. Sailaja, H. Seo and B. S. Reddy, J. Mater. Sci., 2012, 47, 772.
22. R. Saraf, C. Shivakumara, S. Behera, N. Dhananjaya and H. Nagabuhushana, RSC Adv., 2015, 5, 9241.
23. B. R. Judd, Phys. Rev., 1962, 127, 750–761.
24. R. Shukla, S. K. Gupta, V. Grover, V. Nartarajah and A. K. Tyagi, Dalton Trans., 2015, 44, 10628–10635.
25. F. Kang, Y. Zhang and M. Peng, Inorg. Chem., 2015, 54, 1462–1473.
26. H. Li, R. Zhao, Y. Jia, W. Sun, J. Fu, L. Jiang, S. Zhang, R. Pang and C. Li, ACS Appl. Mater. Interfaces, 2014, 6, 3163.
27. H. Yi, F. Li, L. Wu, L. Wu, H. Wang, B. Wang, Y. Zhang, Y. Kong and J. Xu, RSC Adv., 2014, 4, 64244.

---