Syntheses and Luminescence Enhancement of Li1+x(Ta1-zNbz)1-xTixO3:RE3+ (RE=Eu, Er, Tm or Dy) Phosphors

Abstract

We synthesized the phosphors of RE3+-doped Li1+x (Ta1-zNbz)1-xTixO3 (LTNT) with various emission colors by solid-state reaction in air using a conventional electric furnace, where RE=Eu, Er, Tm, or Dy. The optimal host composition with each dopant was determined for the highest photoluminescence (PL) intensity; the relevant chemical formulas were Li1.11Ta0.89Ti0.11O3 (x=0.11 and z=0) with Eu3+, Li1.03 (Ta0.2Nb0.8)0.97Ti0.03O3 (x=0.03 and z=0.8) with Er3+, Li1.08 (Ta0.4Nb0.6)0.92Ti0.08O3 (x=0.08 and z=0.6) with Tm3+, and Li1.14 (Ta0.6Nb0.4)0.86Ti0.14O3 (x=0.14 and z=0.4) with Dy3+. The crystal structures of the phosphors were refined by the XRD-Rietveld method to clarify the relationship between the PL properties and crystal structures. In the LTNT host material, the most effective activator was the Eu3+ ion, with an internal quantum efficiency of 97%. The efficient PL emission is closely related to the coordination environment of Eu3+ in the [Li (Eu) O12] polyhedron of the host LTNT.

Keywords

Oxide phosphors, Photoluminescence, XRD, Rietveld method

Introduction

In the Li₂O–Nb₂O₅–TiO₂ system, the crystals of Li_1+x-yNb_1-x-3yT_ix+4yO₃ (LNT) with 0.05 ≤ x ≤ 0.3 and 0 ≤ y ≤ 0.182 form a superstructure known as the M-phase, which was discovered by Villafuerte-Castrejon et al. [1,2]. Several researchers have investigated the origin of this superstructure, and clarified that it is formed by the periodical insertion of an intergrowth layer into a matrix having a basic trigonal basic structure [3-7]. With the crystals of Li_{1+ x -y}Ta_{1- x -3y}Ti _{x +4y}O₃ (LTT) in the Li₂O–Ta₂O₅–TiO₂ system (0.1≤ x ≤ 0.15 and 0.05≤ y ≤ 0.175), there are a few reports mentioning the relationship between the dielectric properties and compositions [8]. However, to the best of the authors’ knowledge, there have been no reports on the detailed crystal structures. The compositional ranges of the superstructure formation, together with the periods of the intergrowth layer, were different between the crystals of LNT and LTT. Interestingly, the amount of T_i that was required for the superstructure formation was larger for LTT than for LNT [9].

For the application of LNT as a phosphor host material, rare earth (RE³⁺=Eu, Er, Tm, or Dy) doped LNT solid solutions have been prepared by heating in air using a conventional electric furnace and/or millimeter-wave furnace [10-12]. The RE3+-doped LTT phosphors were also successfully synthesized, and their photoluminescence (PL) properties were compared to those of LNT:RE³⁺ [13]. Furthermore, we previously reported on a new red-emitting phosphor in the Li₂O–Nb₂O₅–Ta₂O₅–TiO₂ system, in which we used the quaternary Li_1+x (Ta_1−zNb_z)_1−xTi_xO₃ (LTNT, 0 ≤ x ≤0.25, 0 ≤ z ≤ 1.0) solid solution as the host material [14]. The PL intensity of the LTT:Eu³⁺ phosphor, which was found to be dependent on the concentration of Eu³⁺, showed a high internal quantum efficiency of 84%.

In this work, in order to further improve the PL intensities of the newly developed LTNT:RE³⁺ phosphors, we have clarified the optimal compositions of the host materials for each of the activators RE³⁺, and the relevant crystal structures. We also discussed the close relationship between the PL properties and crystal structures.

Experimental Procedure

The starting materials used for the preparation of LTNT were the reagent-grade chemicals of Li₂CO₃, Nb₂O₅, Ta₂O₅, and TiO₂ (>99.9% grade). Each of the rare earth oxides (Eu₂O₃, Sm₂O₃, Er₂O₃, Tm₂O₃, and Dy2O3>99.9% grade) was doped in the LTNT solid solution. The powder specimens were well mixed, pressed into pellets, and heated in air at 1423 K for 15 h in a conventional electric furnace.

Phase identification was made based on the X-ray powder diffraction (XRPD) data (CuKα), which were obtained on a RINT 2500 device (Rigaku Co., Ltd., Japan) operated at 40 kV and 200 mA. The phase compositions were determined from the XRPD data (CuKα1) collected on another diffractometer in the 2θ range of 15.0-100.0° (X’Pert PRO Alpha-1, PAN alytical B.V., Almelo, the Netherlands) operated at 45 kV and 40 mA. The profile intensity data were investigated by the Rietveld method [15] using a computer program RIETAN-FP [16]. The crystal-structure models were visualized with a computer program VESTA [17]. Distortion parameters for the coordination polyhedra were determined using a computer program IVTON [18]. Excitation and emission spectra were obtained using a fluorescence spectrophotometer (F-7000HITACHI, Japan). Quantum efficiency was measured by a spectral radiometer (MCPD-7000, Otsuka Electronics Co., Ltd., Osaka, Japan).

Results and Discussions

Optimal compositions and PL properties of LTNT: RE3+ phosphors

In our previous paper, we compared the emission spectra between the two-types of phosphors, RE³⁺-doped Li_1.11Nb_0.89Ti_0.11O₀ (LNT with x=0.11 and y=0) and RE³⁺-doped Li_1.11Ta_0.89Ti_0.11O₃ (LTT with x=0.11 and y=0) as shown in Figure 1 [13]. The resulting materials showed various emission colors, ranging from red for Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺, yellow for Li_1.11Ta_0.89Ti_0.11O₃:Dy³⁺, green for Li_1.11Ta_0.89Ti_0.11O₃:Er³⁺, and blue for Li_1.11Ta_0.89Ti_0.11O₃:Tm₃₊. With the Li_1.11Ta_0.89 Ti_0.11O₃ host material, the most appropriate activator for effective emission was the Eu³⁺ ion. On the other hand, when activated with Er³⁺ and/or Tm³⁺, the emission intensities were lower for Li_1.11Ta_0.89Ti_0.11O₃ than for Li_1.11Ta_0.89Ti_0.11O₃. Hence, we concluded that the slight differences in the coordination environment of RE³⁺ between the host materials of Li_1.11Ta_0.89Ti_0.11O₃ and Li_1.11Ta_0.89Ti_0.11O₃ would effectively affect the emission energies of the RE³⁺ ions.

Figure 1: Emission spectra of Li_1.11(Ta/Nb)_0.89 Ti_0.11 O₃:RE³⁺. Dotted line is for LNT:RE³⁺ and solid line is for LTT:RE³⁺ [13].

Figure 2 shows the relationship between PL intensity and the z value of L_1.11(Ta_1-zNb_z)_0.89Ti_0.11:RE³⁺ (RE₂O₃: 2.5 wt%). The emission or excitation wavelengths were measured by monitoring them at maximum wavelengths. The PL intensities of the phosphors were measured at 399 nm for Eu³⁺, at 526 nm for Er³⁺, at 359 nm for Tm³⁺, and at 355 nm for Dy³⁺. We determined the optimal z-values of LTNT (x=0.11) with different RE³⁺ ions; they were z=0 for Eu, z=0.6 for Tm, z=0.8 for Er, and z=0.4 for Dy. We subsequently determined the most suitable quantities of RE³⁺ ions as activators for the host material of Li_1.11Ta_0.89Ti_0.11O₃. Among the phosphors doped with Eu₂O₃ up to 4.5 wt%, the highest PL intensity was observed when doped with 2.5 wt% Eu₂O₃ (Figure 3). For the other phosphors doped with Er₂O₃ up to 3.0 wt%, those doped with Tm₂O₃ up to 3.5 wt%, and those doped with Dy₂O₃ up to 4.0 wt%, the PL intensities were the highest for 1.0 wt% Er₂O₃, 1.0 wt% Tm₂O₃, and 1.0 wt% Dy₂O₃ (Figure 3). Finally, we determined the optimal T_i content in the host materials under the best condition of the z-values and RE₂O₃ contents (Figure 4). As a result, the optimal host composition showing the highest PL intensity was eventually determined for each of the dopants; the relevant chemical formulas were Li_1.11Ta_0.89Ti_0.11O₃ (x=0.11 and z=0) with Eu³⁺, Li_1.03(Ta_0.2Nb_0.8)_0.97Ti_0.03O₃ (x=0.03 and z=0.8) with Er³⁺, Li_1.08(Ta_0.4Nb_0.6)_0.92Ti_0.08O₃ (x=0.08 and z=0.6) with Tm³⁺, and Li_1.14(Ta_0.6Nb_0.4)_0.86Ti_0.14O₃ (x=0.14 and z=0.4) with Dy³⁺. In general, the electronic dipole intensities in the lanthanide 4f–4f transitions exhibit extraordinary sensitivity to the ligand environment, hence the optimal host compositions as determined above must be close to those most appropriate in view of the lattice vibrations and environment of RE³⁺ ions.

Figure 2: Changes of PL intensities with z values of L_1.11(Ta_1-zNb_z)_0.89Ti_0.11:RE³⁺ (RE₂O₃: 2.5 wt%). Phosphors activated with (a) Eu³⁺, (b) Er³⁺, (c) Tm³⁺, and (d) Dy³⁺.

Figure 3: Changes of PL intensities with RE₂O₃ concentrations of Li_1.11Ta_0.89Ti_0.11O₃:RE³⁺. Phosphors activated with (a) Eu³⁺, (b) Er³⁺, (c) Tm³⁺, and (d) Dy³⁺.

Figure 4: Changes of PL intensities with x-values (concentrations of Ti) of (a) Li_1+xTa_1-xTixO₃:Eu³⁺, (b) Li^1+x(Ta_0.2Nb_0.8)_1-xTi_xO₃:Er³⁺, (c) Li_1+x(Ta_0.4Nb_0.6)_1-xTi_xO₃:Tm³⁺, and (d) Li_1+x(Ta_0.6Nb_0.4)_1-xTi_xO₃:Dy³⁺.

In the phosphors emitting various colors, the Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺ phosphor showed a relatively high internal quantum efficiency of 84% as previously reported [14]. Furthermore, we reported that the Sm³⁺ ion can act as an effective sensitizer for Eu³⁺-activated phosphors because of the energy transfer process [17]. In the present experimental stage, we synthesized the Eu³⁺ and Sm³⁺ co-doped phosphor, and characterized the PL property. Figure 5 shows the excitation and emission spectra of Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺, Sm³⁺, in which the contents of Eu₂O₃ and Sm₂O₃ were, respectively, 0.1 wt% and 2.5 wt%. The sharp emission peaks with relatively strong intensities were observed at a wavelength of around 625 nm (⁵D₀-⁷F₂, electric-dipole), when excited by the purple light of 399 nm. The relatively weak emission appeared at wavelengths around 600 nm (⁵D₀-⁷F₁, magnetic-dipole transition in the Eu³⁺ ion) and 700 nm (⁵D₀-⁷F₄, electric-dipole transition in the Eu³⁺ ion). The inner quantum efficiency of the phosphor achieved 97%, which was very close to the theoretical value.

Figure 5: Excitation and emission spectra of Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺,Sm³⁺.

Figure 6 shows the emission and excitation spectra of LTNT phosphors with the most suitable compositions as determined above. Table 1 summarizes the chromaticity coordinates (x, y) of the emitting light. A green emission peak at 526 nm (^4S3/2-^4I15/2) was observed for the Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃:Er³⁺ upon excitation at 551 nm (^4I15/2-^2H11/2). Blue emission was observed at 363 nm (¹D₂-³F₄ transition) for the Li_1.08 (Ta_0.4Nb_0.6)_0.92Ti_0.08O₃:Tm³⁺ upon excitation at 461 nm (³H₆-¹D₂). Yellow emission peaks at around 581 nm (^4F9/2-^6H13/2) were observed for the Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃:Dy³⁺ upon excitation at 356 nm (^6H15/2 -^4M15/2). The PL behavior of the RE³⁺-doped phosphors was not affected by the host material’s structure due to the 4f-4f transitions [13,19].

Figure 6: Emission and excitation spectra of (a) Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃:Er³⁺, (b) Li_1.08 (Ta_0.4Nb_0.6)_0.92Ti_0.08O₃:Tm³⁺, and (c) Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃:Dy³⁺.

Table 1. Optimal compositions and chromaticity of LTNT phosphors.

Morphologies and Crystal Structures of LNTN: RE³⁺ Phosphors

In this chapter, we will discuss the micro-textures and crystal structures of the LTNT:RE³⁺ phosphors. The SEM images in Figure 7 show that the average grain sizes were about 1.5 μm for Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺,Sm³⁺, 5.3 μm for Li_1.08 (Ta_0.4Nb_0.6)_0.92Ti_0.08O₃:Tm³⁺, 8.0 μm for Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃:Er³⁺, and 4.2 μm for Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃:Dy³⁺. The LNT grain’s shape changes from spherical to plate-like when it forms a superstructure [7]. The SEM images indicated that the LTNT:RE³⁺ phosphors had no superstructure because their shapes were not platelike. The difference in sizes was due to the speed of the grain growth owing to the various doping rare-earth ions and their concentrations. The reflection indices in Figure 8 were based on the hexagonal unit cells with a ~ 0.5 nm, c ~ 1.4 nm. Parts of the crystal structures of the host materials were determined, and shown in Figures 9 and 10. We successfully constructed the structural models, in which all of the RE³⁺ ions occupy the Li site. It should be noted that all of these structures were isomorphous with LiNbO₃ (space group R3c), and free from the superstructure peculiar to the M-phase.

Figure 7: SEM images of (a) Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺,Sm³⁺,(b) Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃:Er³⁺, (c) Li_1.08(Ta_0.4Nb_0.6)_0.92Ti_0.08O₃:Tm³⁺, and (d) Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃:Dy³⁺.

Figure 8: XRD patterns of (a) Li_1.11Ta_0.89Ti_0.11O₃:Eu³⁺,Sm³⁺,(b) Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃:Er³⁺, (c) Li_1.08(Ta_0.4Nb_0.6)_0.92Ti_0.08O₃:Tm³⁺, and (d) Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃:Dy³⁺.

Figure 9: Structural models viewed along [11] and structural data of Li_0.901Eu_0.033TaO₃, (Li_0.977Eu_0.023)(Ta_0.89Ti_0.11)O_2.968 and (Li_0.952Eu_0.026)(Ta_0.75Ti_0.25)O_2.901 phosphors.

Figure 10: Structural models viewed along [11] and structural data of (Li_0.991Tm_0.007)(Nb_0.598Ta_0.400)O₃ and (Li_0.990Tm_0.011) (Ta_0.361Nb_0.558Ti_0.081)O_2.970 phosphors.

The addition of relatively small amounts of Ti⁴⁺ effectively increased the PL intensities for Li_1+xTa_1-xTi_xO₃:Eu³⁺ and Li_1+x(Ta_0.4Nb_0.6)_1-xTi_xO₃:Tm³⁺ as shown in Figure 4. Hence, we refined the crystal structures of Li_1+xTa_1-xTi_xO₃:Eu³⁺ with 0 ≤ x ≤ 0.25. Figure 9 shows parts of the refined structural models of Li_0.901Eu_0.033TaO₃ (x=0), (Li_0.977Eu_0.023) (Ta_0.89Ti_0.11)O_2.968 (x=0.11), and (Li_0.952Eu_0.026)(Ta_0.75Ti_0.25)O_2.901 (x=0.25). These chemical compositions were determined as the result of the Rietveld refinement process, and for this reason there are small deviations from those used hitherto. The table included in Figure 9 describes the polyhedral details of Δ (centroid-to-cation distance; eccentricity), VS (sphere volume), and VP (volume of coordination polyhedron). The distortion-parameter values (VS/VP) of [(Li, Eu)O₁₂] polyhedra were close to each other among the three-types of phosphors with x=0, 0.11, and 0.25, as were the VS/VP-values of [(Li, Eu)O₁₂] polyhedra. However, the (Li_0.977Eu_0.023)(Ta_0.89Ti_0.11)O_2.968 phosphor (x=0.11) showed a remarkably large Δ-value (= 0.047 nm) in [(Li, Eu)O₁₂] polyhedron as compared with the two other phosphors with x=0 and 0.25. Because the (Li_0.977Eu_0.023)(Ta_0.89Ti0.11)O_2.968 phosphor showed the highest PL intensity, the displacement of Eu³⁺ position from the centroid of [(Li, Eu)O₁₂] polyhedra must contribute to the highly enhanced intraconfigurational ⁷F₀–⁵D¹ transition.

Figure 10 shows parts of the crystal structures of (Li_0.991Tm_0.007) (Ta_0.400Nb_0.598)O₃ and (Li_0.990Tm_0.011) (Ta_0.361Nb_0.558Ti_0.081)O_2.970. These chemical formulas were determined by the Rietveld method, and hence they show slight deviations from the chemical compositions of Li_1+x(Ta_0.4Nb_0.6)_1-xTi_xO₃:Tm³⁺ with x=0 and 0.08. The Δ-values as well as the VS/VP-values in [(Li, Tm)O₁₂] polyhedra were almost the same between the two phosphors, although the PL intensity was effectively enhanced for x=0.08 than for x=0 as shown in Figure 4c. Thus, the displacement of the Tm³⁺ position from the centroid position of [(Li, Tm)O₁₂] polyhedra would be totally unrelated to the enhancement of PL intensity. Additionally, the displacements of Eu³⁺ and Sm³⁺ positions along the c-axis from the centroid of [(Eu, Sm)O₁₂] polyhedra could be closely related to the enhancement mechanism of the red-light emission through the energy transfer from Sm³⁺ to Eu³⁺ [20,21].

Conclusions

We synthesized LTNT:RE³⁺ (RE=Eu, Er, Tm, and Dy) phosphors with various emission colors by solid-state reaction in air at 1423 K for 15 h using a conventional electric furnace. The phosphors emitted red light for Eu³⁺, green light for Er³⁺, blue light for Tm³⁺, and yellow light for Dy³⁺. The optimal host compositions of the phosphors were Li_1.11Ta_0.89Ti_0.11O₃ with Eu³⁺, Li_1.03 (Ta_0.2Nb_0.8)_0.97Ti_0.03O₃ with Er³⁺, Li_1.08 (Ta_0.4Nb_0.6)_0.92Ti_0.08O₃ with Tm³⁺, and Li_1.14 (Ta_0.6Nb_0.4)_0.86Ti_0.14O₃ with Dy³⁺. These compositions would be close to those most appropriate in view of the lattice vibrations and environment of RE³⁺ ions, since the electronic dipole intensities in the lanthanide 4f–4f transitions generally exhibit extraordinary sensitivity to the ligand environment. For the LTNT host phosphor, the most effective activator was the Eu³⁺ ion. The red-emitting phosphor of Li_1.11Ta_0.89Ti_0.11O₃: Eu³⁺,Sm³⁺ had a high internal quantum efficiency of 97%. A structural study of the phosphors using the XRD-Rietveld method indicated that the high PL intensity of the Eu³⁺-activated phosphor was closely related to the environmental symmetry of the Eu³⁺ ion. The displacement of the Eu³⁺ position from the centroid of [(Li, Eu)O₁₂] polyhedra would effectively contribute to the highly enhanced intraconfigurational ⁷F₀–⁵D₁ transition.

Acknowledgments

This work was partially supported by a Grant-in-Aid for Scientific Research (c) No. 16K06721 (H. N.) by the Japan Society for the Promotion of Science.

References

1. Villafuerte-Castrejon ME, et al. New rutile solid solutions Ti1-4xLixM3xO2:M=Nb, Ta, Sb. J Brit Ceram Soc 1984;83:143-145.
2. Villafuerte-Castrejon ME, et al. Compound and solid-solution formation in the system Li2O-Nb2O5-TiO2. J Solid State Ceram 1987;71:103-108.
3. Smith RI and West AR. Characterization of an incommensurate LiTiNb oxide. Mat Res Bull 1922;27:277-285.
4. Hayashi H, et al. Superstructure in LiTiNb oxides. Forth Ceram Soc 1995;2:391-398.
5. Hayashi H, et al. Preparation of stoichiometric crystalline Li(Nb, Ti)O3 solid solutions by sol-gel processing with metal alkoxides. Key Eng Mater 1999;161-163:501-504.
6. Farber L, et al. Structural study of Li1+x-yNb1-x-3yTix+4yO3 M-Phase solid solutions. J Solid State Chem 2002;166:81-90.
7. Nakano H, et al. Rapid synthesis and structural analysis of Li-Nb-Ti-O solid solutions with superstructure by millimeter-wave heating. J Ceram Soc Jpn 2011;119:808-812.
8. Nakano H, et al. Synthesis and luminescent properties of a new Eu3+-doped Li1+x (Ta1-zNbz)1-xTixO3 red phosphor. J Am Ceram Soc 2012;95:2795-2797.
9. Nakano H, et al. Microstructural comparison between Nb- and Ta-systems in Li1+x-yM1-x-3yTix+4yO3 (M=Nb5+, Ta5+) solid solution with superstructure. J Alloys Comp 2015;618:504-507.
10. Hayashi H and Nakano H. Evaluation and preparation of Li1+x-yM1-x-3yTix+4yO3 solid solution with superstructure as new phosphor. J Alloys Comp 2010;502:360-364.
11. Hayashi H, et al. Microstructure and luminescence of Eu-doped Li1+x-yM1-x-3yTix+4yO3 solid solutions with superstructure. J Ceram Soc Jpn 2010;118:226-230.
12. Nakano H, et al. Rapid synthesis of Eu3+-doped LNT (Li-Nb-Ti-O) phosphor by millimeter-wave heating. Opt Mater 2013;35:2045-2048.
13. Nakano H, et al. Synthesis of New RE3+ Doped Li1+xTa1−xTixO3 (RE: Eu, Sm, Er, Tm, and Dy) Phosphors with Various Emission Colors. Materials 2013;6:2768-2776.
14. Nakano H, et al. Synthesis and luminescent properties of a new Eu3+-doped Li1+x (Ta1−zNbz)1−xTixO3 red phosphor. J Am Ceram Soc 2012;95:2795-2797.
15. Izumi F and Momma K. Three-dimensional visualization in powder diffraction. Solid State Phenom 2007;130:15-20.
16. Brindley GW. The effect of grain or particle Size on x-ray reflections from mixed powders and alloys, considered in relation to the quantitative determination of crystalline substances by x-ray methods. Philos Mag 1945;36:347-369.
17. Momma K and Izumi F. VESTA 3 for three-dimensional visualization of crystal, volumetric andmorphology data. J Appl Crystallogr 2011;44:1272-1276.
18. Balic-Zunic T and Vickovic I. IVTON – a program for the calculation of geometrical aspects of crystal structures and some crystal chemical applications. J Appl Crystallogr 1996;29:305-306.].
19. Nakano H, et al. Synthesis and luminescence enhancement of Eu3+, Sm3+ co-doped Li1.11Ta0.89Ti0.11O3 phosphor. Mater Res Bull 2014;60:766-770.
20. Yan B and Su X-Q. LuVO4:RE3+ (RE=Sm, Eu, Dy, Er) phosphors by in-situ chemical precipitation construction of hybrid precursors. Opt Mater 2007;29:547-551.
21. Lee G-H, et al. Effect of local environment and Sm3+-codoping on the luminescence properties in the Eu3+-doped potassium tungstate phosphor for white LEDS. J Lumin 2008;128:1922-1926.