Photoluminescent and cathodoluminescent performances of Tb³⁺ in Lu³⁺-stabilized gadolinium aluminate garnet solid-solutions of [(Gd_1−xLu_x)_1−yTb_y]₃Al₅O₁₂

Abstract

A series of [(Gd_1−xLu_x)_1−yTb_y]₃Al₅O₁₂ green phosphors have been calcined from their coprecipitated carbonate precursors. Increasing Lu³⁺ incorporation was found to significantly simplify the reaction pathway and lower the temperature of garnet formation, and at the same time leads to contracted cell dimensions and improved theoretical density. Photoluminescent properties, in terms of excitation, emission, concentration quenching, quantum yield, color coordinates of emission, efficiency of excitation absorption, fluorescence decay, and relative intensity of the ⁵D₄ → ⁵F₆/⁵D₄ → ⁵F₅ emissions, are thoroughly investigated against the Lu and Tb contents, temperature of the phosphor synthesis, and electronegativity of the cations in the host lattice. The best luminescent [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG combination exhibits photoluminescence very close to (Y_0.9Tb_0.1)AG and significantly better than (Lu_0.9Tb_0.1)AG. Cathodoluminescence found that the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG phosphor is structurally stable in the range of this study, and exhibits successively higher emission brightness by increasing either the acceleration voltage (up to 6 kV) or beam current (up to 55 μA). The high density green phosphor developed in this work may find lighting, display, and scintillation applications.

---

Introduction

Rare-earth aluminate garnet (Ln₃Al₅O₁₂, LnAG) crystallizes in a bcc crystal structure with 160 atoms in the unit cell (space group: Ia3d), which can be viewed as a framework built up via corner sharing of Al–O tetrahedrons (S₄ point symmetry, 60%) and octahedrons (C_3i symmetry, 40%), with the Ln atoms residing in dodecahedral interstices (D₂ symmetry).¹ LnAG, best known for YAG, is a versatile host lattice for a broad range of optical functionalities, since the Ln site can be facilely substituted with various rare-earth activators. The occurrence and stability of LnAG are, however, heavily dependent on the ionic size of Ln³⁺, and the compound only exists for the relatively small ions of Gd³⁺–Lu³⁺ in the lanthanide family (including Y³⁺). Gd is the boundary for LnAG to be formed, but GdAG (further abridged as GAG) is metastable and decomposes into Al₂O₃ and GdAlO₃ (Gd₃Al₅O₁₂ → Al₂O₃ + 3GdAlO₃) when the temperature is above ∼1300 °C.^2,3 GAG can, nonetheless, be stabilized via two strategies, that is, doping the Gd site with a smaller lanthanide to reduce the average ionic size of this site so that the Ln³⁺ ions can enter the dodecahedrons or modifying the Al site with a suitably larger trivalent ion (such as Ga³⁺)⁴ to enlarge the interstices so that the Gd³⁺ ions can be accommodated therein. We previously demonstrated that replacing Gd³⁺ with ∼10 at% of Lu³⁺, 22 at% of Y³⁺ or 50 at% of Tb³⁺ can yield thermodynamically stable garnet solid-solutions, with the bandgap energy and optical property tunable by dopant concentration.^3,5

Compared with YAG, GAG may hold a number of merits for optical applications: (1) GAG has a significantly higher theoretical density (∼5.97 g cm⁻³) than YAG (∼4.55 g cm⁻³) and the atomic weight of Gd (∼157, close to the 175 of Lu) is much higher than Y (∼89), and thus GAG is more desired for scintillation purposes. GAG-based scintillators that show superior light yields and fast emission decay were recently reported;^6–9 (2) the lower electronegativity of Gd³⁺ (χ = 1.20) than Y³⁺ (χ = 1.22) may produce new features and improved intensity of emission. For example, the (Gd,Lu)AG : Eu red phosphor shows better emission than YAG : Eu under identical powder synthesis,^10,11 and the (Gd,Lu)AG : Ce yellow phosphor has more red-component than YAG : Ce in its emission spectrum owing to enhanced crystal field splitting of the Ce³⁺ 5d energy level,¹² which is needed for warm-white LED lighting; (3) the intrinsic ⁸S_7/2 → ⁶I_J intra-4f⁷ transition of Gd³⁺ (∼275 nm) can be utilized as an efficient excitation source for certain activators, and greatly improved luminescence is obtainable with the energy transfer from Gd³⁺. We recently showed that the (Gd,Lu)AG : Dy white phosphor emits twice stronger through exciting Gd³⁺ at 275 nm rather than directly exciting Dy³⁺ at 352 nm (the ⁶H_15/2 → ⁴I_11/2 + ⁴M_15/2 + ⁶P_7/2 transition of Dy³⁺, the strongest intra-4f⁹ transition), and the emission intensity under 275 nm excitation reaches three and six times those of YAG : Dy and LuAG : Dy under 352 nm excitation, respectively.¹³

YAG : Tb is a green phosphor widely studied for applications in cathode ray tubes, scintillation, and flat panel displays such as field emission display (FED), vacuum fluorescent display (VFD), and electroluminescent (EL) display, since it is thermally and chemically stable and resists saturation under high current excitation.^14,15 We synthesized in this work a series of [(Gd_1−xLu_x)_1−yTb_y]₃AG solid solutions (x = 0.1–0.5, y = 0–0.15) with Lu³⁺ as a lattice stabilizer, and the luminescent properties were thoroughly investigated in terms of the Lu³⁺/Tb³⁺ content, processing temperature, and fluorescence decay. The emission properties were also compared with those of YAG : Tb and LuAG : Tb under identical conditions of phosphor processing. In the following sections, we report the synthesis via carbonate coprecipitation, phase evolution, structure characterization, and photoluminescent and cathodoluminescent properties of the [(Gd_1−xLu_x)_1−yTb_y]₃AG green phosphors.

Experimental

The rare earth sources for powder synthesis are RE₂O₃ (RE = Gd, Lu, Y) and Tb₄O₇, all 99.99% pure products from Huizhou Ruier Rare Chemical Hi-Tech Co. Ltd (Huizhou, China). The aluminum source is >99% pure alum (NH₄Al(SO₄)₂·12H₂O) from Zhenxing Chemical Reagent Factory (Shanghai, China). All the other reagents are of analytical grade products from Shenyang Chemical Reagent Factory (Shenyang, China). Nitrate solution of the rare-earth was prepared by dissolving the corresponding oxide in concentrated HNO₃, followed by evaporation to dryness at 100 °C to remove superfluous HNO₃ and then a final dilution with distilled water to 0.3 mol L⁻¹.

The precursor for phosphor oxide was made via carbonate coprecipitation with ammonium hydrogen bicarbonate (NH₄HCO₃, AHC) as the precipitant, which has the advantages of stoichiometric precipitation of the constituent cations and low aggregation of the final oxide powders.^10,16,17 Aqueous salt solutions for precipitation were made from the nitrate solutions and alum according to the formula of [(Gd_1−xLu_x)_1−yTb_y]₃Al₅O₁₂ (x = 0.1–0.5, y = 0–0.15). Carbonate precursors were precipitated by dropwise adding 200 mL of a 0.125 mol L⁻¹ (for Al³⁺) mixed solution into 320 mL of a 1.5 mol L⁻¹ AHC solution under mild stirring at room temperature. In all the cases, the total concentration of Gd³⁺, Lu³⁺ and Tb³⁺ was kept constant at 0.075 mol L⁻¹ so that the (Gd + Lu + Tb) : Al atomic ratio would be the 3 : 5 of the garnet formula. The precipitate was homogenized for 30 min after the completion of precipitation, centrifuged, and washed with distilled water and alcohol to remove by-products. The resultant wet precipitate was dried in air at 100 °C for 24 h, lightly crushed, and finally calcined to produce oxides. To prevent Tb³⁺ oxidization, the carbonate precursors were first decomposed at 600 °C for 4 h in air, and then calcined at higher temperatures (holding time: 4 h) in a Ar (95 vol%)/H₂ (5 vol%) gas mixture flowing at 200 mL min⁻¹, (Y_0.9Tb_0.1)AG and (Lu_0.9Tb_0.1)AG (x = 1.0) samples were similarly synthesized for comparison.

Phase identification was performed via X-ray diffractometry (XRD, Model PW3040/60, Panalytical B.V., Almelo, The Netherlands) using nickel-filtered Cu-Kα radiation and a scanning speed of 4° 2θ min⁻¹. Morphology of the product was observed by field-emission scanning electron microscopy (FE-SEM, Model JSM-7001F, JEOL, Tokyo, Japan). Specific surface area of the oxide powder was analyzed on an automatic analyzer (Model TriStar II 3020, Micromeritics Instrument Corp., Norcross, GA) using the Brunauer–Emmett–Teller (BET) method via nitrogen adsorption at 77 K. Photoluminescence, fluorescence decay, and quantum yield of the phosphor were measured at room temperature using an FP-6500 fluorospectrophotometer (JASCO, Tokyo) equipped with a 60 mm diameter integrating sphere (Model ISF-513, JASCO) and a 150 W Xe-lamp for excitation. Monochromatization of the excitation and emission lights was achieved with a Rowland concave grating (1800 grooves mm⁻¹). Optical measurements were conducted under identical conditions for all the samples, with slit widths of 5 nm for both excitation and emission. Spectral responses of the equipment were corrected in the range 220–850 nm with a Rhodamine-B solution (5.5 g L⁻¹ in ethylene glycol) and a standard light source unit (ECS-333, JASCO) as references. Cathodoluminescence of the phosphor powder was measured by a spectroradiometer (HS-1000, Photal Otsuka Electronics, Osaka, Japan) in a vacuum chamber with a base pressure of 10⁻⁶ Pa at room temperature, after being tightly pressed on an Al plate of 1 × 1 cm². The electron beam covers a circular area of 0.76 cm in diameter (∼0.45 cm²), from which luminescence signals were collected in the 330–900 nm region with a CCD camera. The brightness of luminescence (cd m⁻²) was then derived from the emission integrated over the wavelength region of measurement with a built-in software of the system.

Results and discussion

Fig. 1a displays temperature-course phase evolution of the [(Gd_1−xLu_x)_0.9Tb_0.1]AG precursor for x = 0.10, from which it is seen that the product remains essentially amorphous up to ∼800 °C, followed by crystallization at ∼900 °C to yield a mixture of the Ln₄Al₂O₉ monoclinic (LnAM, Ln = Gd, Lu, Tb), LnAlO₃ perovskite (LnAP), and Ln₂O₃ phases. The LnAM phase was formed via the reaction between Ln₂O₃ and amorphous Al₂O₃ according to 2Ln₂O₃ + Al₂O₃ → Ln₄Al₂O₉, while the LnAP phase may have been generated via Ln₂O₃ + Al₂O₃ → 2LnAlO₃ and/or Ln₄Al₂O₉ + Al₂O₃ → 4LnAlO₃. The latter two reactions occur simultaneously in the 900 to 1000 °C range, since both the LnAM and Ln₂O₃ phases were completely consumed to yield significantly more LnAP in the 1000 °C product. Further reaction of LnAP with amorphous Al₂O₃ (3LnAlO₃ + Al₂O₃ → Ln₃Al₅O₃) above 1000 °C produces the LnAG phase, and only trace LnAP is detectable at 1300 °C. Phase pure LnAG was resulted by calcination at 1500 °C, which again confirms that 10 at% of Lu³⁺ can fully stabilize the garnet lattice.^3,5,9–13 The lattice stabilization can be understood from the fact that the [(Gd_0.9Lu_0.1)_0.9Tb_0.1]³⁺ combination has an average ionic radius (∼0.1045 nm) very close to the size of Tb³⁺ in TbAG (Gd³⁺, Lu³⁺ and Tb³⁺ have ionic radii of 0.1053 and 0.0977 and 0.1040 nm for CN = 8, respectively),¹⁸ while TbAG is known as a thermodynamically stable garnet that can be further doped with properly larger ions (such as Ce³⁺) for optical functionality. Comparing Fig. 1a–c reveals that the reaction pathway becomes much simpler and LnAG crystallizes at a significantly lower temperature with increasing Lu³⁺ addition. Aside from the stabilization effects of Lu³⁺, this is primarily owing to improved cation homogeneity of the precursor, since Lu³⁺, compared with Gd³⁺, is much closer to Al³⁺ in solution chemistry,¹⁹ which promotes instantaneous coprecipitation of the constituent cations. The lowest temperature to crystallize phase-pure [(Gd_1−xLu_x)_1−yTb_y]₃AG is 1000 °C in this work, for the samples of x = 0.3–0.5. Fig. 1d compares XRD patterns of the products calcined at 1500 °C. No impurity was identified along with the garnet phase in each case, and the diffractions can be well indexed with those of the cubic structured GAG (space group: Ia3d, JCPDS: 1-73-1371). Steady shifting of the XRD peaks towards higher diffraction angles was observed along with increasing Lu³⁺ incorporation due to the smaller ionic radius of Lu³⁺. Fig. 2 shows the calculated cell parameter, from which it is clearly seen that the value almost linearly decreases towards a higher Lu content (x) and follows Vegard's law, implying the already formation of homogeneous solid solution. The theoretical density linearly increases with increasing x, and is higher than that (5.97 g cm⁻³) of the already high GAG in each case. The stabilized garnet structure and the high effective atomic number and theoretical density may permit the [(Gd_1−xLu_x)_0.9Tb_0.1]AG solid solutions to be better scintillation materials than YAG : Tb (∼4.55 g cm⁻³).

Fig. 1 XRD patterns of the [(Gd_1−xLu_x)_0.9Tb_0.1]AG precursors calcined at various temperatures, with (a) x = 0.1, (b) x = 0.2, and (c) x = 0.3, (d) is a comparison of the XRD patterns of the products calcined at 1500 °C. The calcination temperature and Lu content (the x value) are indicated in the figures. Letters G, P, M and R represent LnAG garnet (Ln = Gd, Lu, Tb), LnAP perovskite, LnAM monoclinic, and Ln₂O₃ phases, respectively. All the unlabeled peaks belong to the LnAG phase.

Fig. 2 Lattice constant and theoretical density of the [(Gd_1−xLu_x)_1−yTb_0.1]AG solid solution calcined at 1500 °C, as a function of the Lu content.

Fig. 3 shows typical FE-SEM morphologies of the green phosphors, with the composition [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG for example. It is seen that the originally rounded oxide crystallites/particles of the 1000 °C product undergo considerable growth at the higher temperatures up to 1500 °C, but are generally better dispersed than those of the solid-reaction product. BET analysis found specific surface areas of 20.15, 9.84, 3.85, and 0.19 m² g⁻¹ for the 1000, 1150, 1300, and 1500 °C powders, from which the average particle sizes were estimated to be ∼47 nm, 97 nm, 248 nm, and 5.02 μm, respectively. It can thus be said that significant coarsening takes place when the calcination temperature is above ∼1300 °C.

Fig. 3 FE-SEM micrographs showing morphologies of the [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG powders calcined at 1000 (a), 1300 (b), and 1500 °C (c). The inset in (c) is the appearance of green emission under 254 nm excitation from a hand-held UV lamp.

Fig. 4 shows photoluminescence excitation (PLE) and photoluminescence (PL) behaviors of the [(Gd_0.8Lu_0.2)_1−yTb_y]AG phosphors synthesized at 1150 °C. The excitation spectrum obtained by monitoring the green emission at ∼545 nm is composed of three groups of bands located at ∼227 (E^1–2₃ level, spin allowed), 275 (E^1–3₂, spin allowed, the strongest), and 325 nm (E₁, spin forbidden) in each case, which are arising from the well-documented 4f⁸ → 4f⁷5d¹ inter-configurational Tb³⁺ transitions.²⁰ It should be noted that the ⁸S_7/2 → ⁶I_J Gd³⁺ transition at ∼275 nm completely overlaps the f → d transition of Tb³⁺ and is thus not separable, though it is clearly observed at the same wavelength in the UV/Vis absorption spectra of the (Gd_1−xLu_x)AG hosts (x = 0–0.5)³ and appears for some other phosphor systems activated with Tb³⁺, such as Sr₂GdF₇ : Tb/Eu²¹ and Gd₂O₃ : Tb.²² The overlapping suggests the likelihood of Gd³⁺ → Tb³⁺ energy transfer, since the ⁶I_J state of Gd³⁺ lies higher than the ⁵D_3,4 emission states of Tb³⁺ in the energy diagram of excited states for Ln³⁺.^23–25 The energy transfer was expected to be able to improve Tb³⁺ emission, its contribution, however, cannot be quantitatively analyzed due to the complete overlapping of Gd³⁺ and Tb³⁺ excitations. The phosphors all exhibit four groups of emission bands at ∼490, 545, 589, and 623 nm under 275 nm excitation, with the green emission at 545 nm being the strongest. The bands are known to be associated with transitions from the ⁵D₄ excited state to the ⁷F_J (J = 3–6) ground states as labeled in the figure. Emission from the higher ⁵D₃ level is observed as a very weak band for the y = 0.025 sample, with an intensity only ∼1.5% of the 545 nm emission. The ⁵D₃ emission steadily lowers with increasing Tb³⁺ incorporation and is almost completely quenched at y = 0.10, though the data are not included for overall clarity of the figure. This can be explained by cross-relaxation via resonance between the excited and ground states of two Tb³⁺ ions, that is, populating the ⁵D₄ level by quenching the ⁵D₃ level via Tb³⁺ (⁵D₃) + Tb³⁺ (⁷F₀) → Tb³⁺ (⁵D₄) + Tb³⁺ (⁷F₆).²⁶ Emission from the ⁵D₃ state was found as a relatively strong band for YAG doped with 1 at% of Tb³⁺ (ref. 14) but not with 5 at% of Tb³⁺,¹⁵ and 1 at% is generally accepted as the up-limit for the ⁵D₃ emission to appear in many hosts. It is seen that Tb³⁺ content does not bring about any appreciable change to the peak position of each of the excitation and emission bands, but the emission intensity steadily improves with increasing Tb³⁺ incorporation until x = 0.10 and then decreases (Fig. 4, the inset). The optimal Tb³⁺ content of ∼10 at% (x = 0.10) found in this work is almost identical to that observed from the YAG : Tb green phosphor system.²⁷

Fig. 4 PLE and PL spectra of the [(Gd_0.8Lu_0.2)_1−yTb_y]AG phosphors calcined at 1150 °C. The PLE spectra were taken by monitoring the 545 nm green emission, and the PL spectra were measured under UV excitation at 275 nm. The inset shows relative intensity of the 545 nm emission, normalized to the y = 0.025 sample, as a function of the Tb³⁺ content.

The mutual interaction type of luminescence quenching in solid-state phosphors can be analyzed with the following equation to conclude the quenching mechanism:^28,29

log(I/c) = (−s/d)log(c) + log f (1)

where I is the emission intensity, c the activator content, d the sample dimension (d = 3 for energy transfer among the activators inside particles), f a constant independent of activator concentration, and s the index of electric multipole. The s values of 6, 8, and 10 are for dipole–dipole, dipole–quadrupole, and quadrupole–quadrupole electric interactions, respectively, whereas s = 3 corresponds to exchange interaction. The log(I/c) − log(c) plot is displayed in Fig. 5 for the [(Gd_0.8Lu_0.2)_1−yTb_y]AG samples, from which a slope (−s/3) of −0.78 ± 0.11 was derived via linear fitting. The s value is thus around 3, indicating that the observed luminescence quenching may have been dominantly resulted from exchange interactions for the energy transfer among Tb³⁺, possibly via phonon-assisted non-resonant interactions.³⁰

Fig. 5 The relationship between log(I/c) and log(c) for the [(Gd_0.8Lu_0.2)_1−yTb_y]AG phosphors calcined at 1150 °C.

Ozawa³¹ proposed that, for luminescence quenching via exchange interactions, the quenching concentration can be estimated to be 1/(1 + z) with the z value derived from the equation I = bc(1 − c)^z, where I is the emission intensity, c the activator content, b an independent constant, and z the nearest lattice sites for the activator. The log(I/c) − log(1 − c) plot is presented in Fig. 6 for the green phosphors, from which the z value was analyzed to be ∼10.27 ± 0.59 via linear fitting. The optimal Tb³⁺ concentration is thus ∼8.9%, close to the experimentally found value of ∼10 at%. This may further confirm that the observed concentration quenching of luminescence may largely be due to exchange interactions.

Fig. 6 The relationship between log(I/c) and log(1 − c) for the [(Gd_0.8Lu_0.2)_1−yTb_y]AG phosphors calcined at 1150 °C.

The effects of synthesis temperature on PLE and PL properties are studied in Fig. 7 with [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG as an example, since the low crystallization temperature of this composition (1000 °C) allows an investigation in a wider temperature range. It is seen that, for a fixed host composition, raising the processing temperature from 1000 to 1500 °C does not bring about any appreciable change to the peak position of each of the excitation and emission bands. Since the exposed Tb³⁺ 5d¹ electron readily interacts with the surrounding ligands and the 5d¹ energy level is subjected to crystal field splitting, the excitation behaviors thus confirm that garnet solid solution has been formed in high phase purity, conforming to the results of XRD (Fig. 1). The emission intensity steadily improves towards a higher synthesis temperature, particularly above 1150 °C, following the tendency observed from the excitation spectra. The inset in Fig. 7 shows that the 1500 °C sample has an intensity 182% that of the 1000 °C one for the ⁵D₄ → ⁷F₅ green emission, and the improvement is primarily owing to enhanced crystallization and increased particle/crystallite size as revealed by the results of XRD (Fig. 1) and morphology observations (Fig. 3).

Fig. 7 PLE and PL spectra for the [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG phosphors processed at different temperatures. The inset is the relative intensity of the 545 nm emission, as a function of the synthesis temperature, where the emission intensity is normalized to that of the 1000 °C sample.

At the optimal Tb³⁺ content of 10 at% and under the same synthesis temperature of 1500 °C, photoluminescent properties of the [(Gd_1−xLu_x)_0.9Tb_0.1]AG green phosphors are compared with (Y_0.9Tb_0.1)AG and (Lu_0.9Tb_0.1)AG (x = 1.0) in Fig. 8. The PLE spectrum obtained by monitoring the ∼545 nm green emission is composed of the three groups of 4f⁸ → 4f⁷5d¹ transition and dominated by the band at ∼275 nm in each case, as found in Fig. 4 and 7. The exact peak position of the strongest excitation and the span of the total f → d transition are, however, clearly dependent on the composition of the host lattice owing to different crystal-field splitting of the Tb³⁺ 5d energy level. The extent of splitting is mainly affected by coordination geometry, following the order of octahedral > cubic > dodecahedral > tricapped trigonal prisms and cuboctahedral,³² and also electronegativity of the cations in the host.³³ As the Tb³⁺ activator assumes the same dodecahedral coordination in all the garnets discussed in this work, electronegativity (χ) thus plays an essential role. Lu and Y have χ values of 1.27 and 1.22, respectively, while the (Gd_1−xLu_x) combinations were calculated to have increasing average χ of 1.207, 1.214, 1.221, 1.228, and 1.235 with increasing x from 0.1 to 0.5 (0.1 interval). The garnet hosts can thus be placed in the order LuAG > (Gd_0.5Lu_0.5)AG > (Gd_0.6Lu_0.4)AG > YAG ≈ (Gd_0.7Lu_0.3)AG > (Gd_0.8Lu_0.2)AG > (Gd_0.9Lu_0.1)AG following descending χ. The main f → d excitation indeed tends to occur at a longer wavelength and the whole f → d transition tends to span a wider wavelength region with decreasing χ (Fig. 8a and Table 1), which are conforming to the theoretical prediction that a lower electronegativity would enhance crystal field splitting of the 5d energy level.³³ The peak position of each of the ⁵D₄ → ⁷F_J (J = 6–3) emission, however, is not affected by the host lattice (Fig. 8b), since the emission only involves transitions within the shielded 4f subshell of Tb³⁺. Intensity of the ⁵D₄ → ⁷F₅ green emission, nonetheless, significantly varies with the host and is the strongest for YAG, which is ∼10% higher than that of LuAG. The inferior emission of LuAG : Tb could be due to two reasons: (1) larger lattice distortion due to the bigger size mismatch between Lu³⁺ and Tb³⁺, and (2) the existence of more anti-site defects (ADs). Though the Lu³⁺ ions are assumed to solely reside at the Ln sites of LnAG and occupy the dodecahedral interstices, it is found in practice that they may partially take Al³⁺ sites to form the ADs of Lu_Al and Al_Lu.^34–38 Such defects commonly exist in rare-earth aluminate garnets, but are more readily formed in LuAG since the ionic size of Lu³⁺ is the closest to that of Al³⁺ in the whole lanthanide family (Y³⁺ close to Ho³⁺ in size).¹⁸ ADs are known to hamper energy delivery to the activator, leading to deteriorated luminescence and also longer fluorescence lifetime.³⁴ Though the lifetime determined for the 545 nm green emission does not vary significantly among the phosphors, LuAG : Tb does have the largest value in this work (Table 1). Among the [(Gd_1−xLu_x)_0.9Tb_0.1]AG green phosphors, the x = 0.2 sample is the most luminescent and has an emission intensity almost identical to that of YAG : Tb (Fig. 8b and Table 1). The emission tends to deteriorate with increasing Lu incorporation, which can also be explained with the above two reasons. That is, enhanced lattice distortion owing to the large size mismatch between Gd³⁺ and Lu³⁺ and higher probability of ADs formation at a higher Lu content. It is also seen from Table 1 that the fluorescence lifetime tends to be longer with more Lu in the lattice. The better luminescence of the x = 0.2 than x = 0.1 sample could be due to its higher crystallinity, since it has already crystallized as a pure garnet at the low temperature of 1150 °C while the latter needs ∼1500 °C to be formed (Fig. 1). Intensity ratio of the 490 nm (⁵D₄ → ⁷F₆) to 545 nm (⁵D₄ → ⁷F₅) emissions (I₄₉₀/I₅₄₅) was observed to vary with the host, being the smallest for (Gd_0.9Lu_0.1)AG (∼0.44) and the largest for LuAG (∼0.6). The value tends to increase with decreasing cell parameter of the garnet host, which is understandable from the enhanced cross-relaxation via Tb³⁺ (⁵D₃) + Tb³⁺ (⁷F₀) → Tb³⁺ (⁵D₄) + Tb³⁺ (⁷F₆).

Fig. 8 A comparison of the PLE and PL behaviors of the [(Gd_1−xLu_x)_0.9Tb_0.1]AG, (Y_0.9Tb_0.1)AG, and (Lu_0.9Tb_0.1)AG green phosphors calcined at 1500 °C.

Table 1 A summary of the properties of the garnet phosphors discussed in this work

	Lu	(Gd0.5Lu0.5)	(Gd0.6Lu0.4)	(Gd0.7Lu0.3)	Y	(Gd0.8Lu0.2)	(Gd0.9Lu0.1)
Electronegativity (χ)	1.27	1.235	1.228	1.221	1.22	1.214	1.207
Main 4f^8 → 4f^75d^1 (nm)	270	274	274	275	275	276	277
4f^8 → 4f^75d^1 span (nm)	91	96	96	97	96	98	99
I545 (%)	91.4	72	86.9	93.4	100	99.3	95.2
I490/I545	0.6	0.49	0.47	0.46	0.49	0.44	0.44
Color coordinates	(0.349, 0.562)	(0.347, 0.563)	(0.348, 0.563)	(0.348, 0.563)	(0.349, 0.563)	(0.348, 0.564)	(0.349, 0.564)
Lifetime (ms, ±0.01)	3.48	3.39	3.44	3.31	3.40	3.29	3.27
εin (%)	82.4	74.3	78.9	81.6	85.4	84.3	83.4
εex (%)	59.6	54.8	58.5	62.7	67.6	65.8	65.6
η (%)	72.3	73.7	74.1	76.8	79.1	78.0	78.7

---

The Commission International de L'Eclairage (CIE) chromaticity coordinates (x, y) were analyzed to be around (0.35, 0.56) for all the garnet-based green phosphors discussed in this work (Table 1), typical of a vivid green color in the CIE chromaticity diagram, as also seen from the digital picture (Fig. 3c). The correlated color temperature (CCT) was estimated to be ∼5200 K for the green phosphors, with the following equations proposed by McCamy:³⁹

T = −437n³ + 3601n² − 6861n + 5514.31, (2)

and

n = (x − 0.332)/(y − 0.1858). (3)

Quantum yield of the phosphor was analyzed by comparing its response to the excitation with a diffusive white standard, as shown in Fig. 9 for the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG green phosphor (x = 0.2). The external quantum efficiency (ε_ex, the total number of emitted photons divided by the total number of excitation photons) and the internal quantum efficiency (ε_in, the total number of emitted photons divided by the number of excitation photons absorbed by the sample) were obtained from the following equations using the software built in the spectrophotometer:

where E(λ)/hν, R(λ)/hν, and P(λ)/hν may present the number of photons in the excitation, reflectance, and emission spectra of the samples, respectively. The reflection spectrum of a Spectralon white standard was used for calibration. The percentage of excitation absorption (η) can be derived as η = (ε_ex/ε_in) × 100%. The data of analysis are tabulated in Table 1, from which it is seen that the LuAG : Tb phosphor has the lowest efficiency of excitation absorption, leading to a relatively low ε_ex though the ε_in is yet high. The η value successively decreases with increasing Lu content for the (Gd_1−xLu_x)AG : Tb solid solutions, with those of the x = 0.1 and 0.2 samples close to that of YAG : Tb. Internal quantum efficiencies of over 80% were observed for LuAG : Tb, YAG : Tb, and the x = 0.1–0.3 samples, with that of YAG : Tb being the highest (85.4%). The decreasing ε_in of (Gd_1−xLu_x)AG : Tb at a higher Lu content may largely be due to increased lattice distortion. The (Gd_1−xLu_x)AG : Tb green phosphors of x = 0.1 and 0.2 have ε_ex values close to that of YAG : Tb.

Fig. 9 Determination of quantum yield (QY) for the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG green phosphor (x = 0.2). The secondary light of the 275 nm excitation is automatically extracted by the QY calculation software.

The kinetics of fluorescence decay is frequently analyzed with the following bi-exponential equation:^40,41

I(t) = A₁ exp(−t/τ₁) + A₂ exp(−t/τ₂) + B (6)

where I(t) is the instantaneous luminescence intensity, A₁, A₂ and B are constants, t is the decay time, and τ₁ and τ₂ are the rapid and slow components of the lifetime, respectively. The A₁, A₂, τ₁, and τ₂ values can be obtained by fitting the decay kinetics with the above equation, and the effective fluorescence lifetime (τ*) can be derived to be:

τ* = (A₁τ₁² + A₂τ₂²)/(A₁τ₁ + A₂τ₂). (7)

Practical fitting found that the fluorescence decay follows the single exponential equation of I = A exp(−t/τ) + B in each case, as shown in Fig. 10 for the 545 nm emission of the [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG phosphor calcined at 1500 °C. This indicates the absence of slow component in the lifetime and the decay is operated by one single mechanism. The inset in Fig. 10 shows the lifetime as a function of the temperature of phosphor synthesis. A significant lifetime shortening from 6.83 ± 0.03 to 3.31 ± 0.01 ms was observed from 1000 to 1500 °C, and this could be attributed to the elimination of lattice defects along with crystal perfection. Lattice defects (such as ADs) are known to be able to temporarily arrest the back-jumping electrons and thus lead to a longer fluorescence lifetime. In addition, particle size may influence the rate of radiative transition and thus the fluorescence lifetime via affecting the effective refractive index of the phosphor particle.⁴² The radiative lifetime can be related to the refractive index using the equation:⁴³

where f(ED) is the oscillator strength for dipole transition, λ₀ is the wavelength in vacuum, and n_eff is the effective refractive index and n_eff = n_cx + (1 − x)n_med, where x is the filling factor, that is, the fraction of space occupied by particle in the surrounding medium (air in this work), and n_c and n_med are the refractive indices of the bulk material and surrounding medium, respectively. n_eff roughly equals n_med for extremely small particles, since the x value approaches 0. For the intermediately sized particles of this work, n_eff increases with the increased particle size attained at a higher calcination temperature (Fig. 3) and thus contributes to lifetime shortening. A shortened lifetime at a bigger particle size was also previously observed for other phosphors.^11,13,29

Fig. 10 Fluorescence decay curve for the 545 nm emission of [(Gd_0.7Lu_0.3)_0.9Tb_0.1]AG calcined at 1500 °C. The inset is the lifetime as a function of calcination temperature, where the standard deviation is given in percentage.

Fig. 11 shows cathodoluminescence (CL) properties for the two best luminescent phosphors of (Y_0.9Tb_0.1)AG and [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG synthesized at 1500 °C. The emission spectra (Fig. 11a) taken for [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG under the constant electron-beam current of 50 μA are similar to the PL emission spectra exhibited in Fig. 8b. The successively improved emission intensity along with increasing acceleration voltage and the almost constant color coordinates indicate that the phosphor is stable in the studied range. For cathodoluminescence, the activator ions are excited by the plasma produced by the incident electrons. Electron energy increases with increasing acceleration voltage, allowing the electrons to penetrate deeper into the phosphor particles, which leads to increased volume of electron–solid interactions and thus improved CL brightness of both the phosphors (Fig. 11c). On the other hand, more plasma will be generated at a higher beam current, and hence enhanced emission brightness was observed for the two phosphors under the fixed acceleration voltage of 5 kV (Fig. 11b). Campo et al.⁴⁴ very recently suggested that high irradiation settings may yield an increased affected area, ultimately motivated by electric charge injection, and thus affects CL intensity. This, however, can be excluded in this work since both the beam current and acceleration voltage adopted herein are mild for garnet based phosphors^14,15 and the luminescence signals are solely collected from the fixed circular area of electron irradiation. The emission intensity of [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG is lower than its YAG counterpart either under fixed beam current or acceleration voltage. At 5 kV and 50 μA, for example, the emission brightness of [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG is ∼60% of that of (Y_0.9Tb_0.1)AG. This can be understood from the excitation depth of the incident electrons. The electron penetration depth can be correlated to acceleration voltage and material properties via the equation L[Å] = 250(A/ρ)(E/Z^1/2)ⁿ, where n = 1.2/(1 − 0.29 log₁₀^Z), A is the atomic or molecular weight of the compound, ρ the bulk density, Z the atomic number per molecule in the material, and E the acceleration voltage (kV).^45–47 The results of calculation are displayed in Fig. 11d for the two phosphors. In view that the penetration depth of [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG (∼71.2 nm) is only one third of that (∼210.2 nm) of (Y_0.9Tb_0.1)AG under 5 kV, the emission efficiency of the former can be said to be pretty high. Moreover, the higher self-absorption of (Y_0.9Tb_0.1)AG than [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG, arising from the smaller atomic weight of Y, was thought to be more relevant to the stronger emission of the former. In addition, Cherepy et al. concluded from their work on TbAG and GAG-based garnet scintillators that the efficient energy migration from the Tb³⁺ and Gd³⁺ sensitizers in the host lattice to the activators under high energy excitations may contribute to extraordinarily high light-yields.⁶

Fig. 11 Cathodoluminescence spectra taken for the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG green phosphor under the fixed irradiation beam current of 50 μA (a), the brightness of emission under varying irradiation current (b) and acceleration voltage (c), and the calculated penetration depth as a function of the acceleration voltage (d). Both the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG and (Y_0.9Tb_0.1)AG phosphors are synthesized at 1500 °C.

Conclusions

A series of [(Gd_1−xLu_x)_1−yTb_y]₃Al₅O₁₂ green phosphors have been synthesized in this work via carbonate coprecipitation, and 10 at% of Lu (x = 0.1) was found to be able to fully stabilize the garnet lattice against thermal decomposition. Increasing Lu³⁺ incorporation significantly simplifies the reaction pathway and lowers the temperature of garnet formation, and at the same time leads to contracted cell dimension and improved theoretical density of the material. The optimal content of Tb³⁺ in the garnet lattice was determined to be ∼10 at% (y = 0.1), and concentration quenching of luminescence was suggested to be resulted from exchange interactions. Though photoluminescence tends to deteriorate towards a higher Lu content, the best composition of [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG exhibits emission properties very close to (Y_0.9Lu_0.1)AG and significantly better than (Lu_0.9Lu_0.1)AG. Kinetics analysis found that the 545 nm green emission decays in a single exponential manner, and successively shorter lifetime was found for the phosphor synthesized at a higher temperature. Cathodoluminescence analysis found that the [(Gd_0.8Lu_0.2)_0.9Tb_0.1]AG phosphor is structurally stable in the range of this study and exhibits successively higher emission brightness by increasing either the acceleration voltage (up to 6 kV) or beam current (up to 55 μA). The high density green phosphor developed in this work may thus find applications in lighting, display, and scintillation areas.

Acknowledgements

This work was supported in part by the National Natural Science Foundation of China (Grants No. 51172038 and 51402125), the research award fund for young and middle-aged scientist in Shandong Province, and the Research Fund for the Doctoral Program of University of Jinan (Grants XBS1447). The authors are indebted to Dr Xiaojun Wang of the National Institute for Materials Science (NIMS) for cathodoluminescence measurements.

References and notes

1. Y.-N. Xu and W. Y. Ching, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59, 10530.
2. T. Shishido, K. Okamura and S. Yajima, J. Am. Ceram. Soc., 1978, 61, 373.
3. J. K. Li, J.-G. Li, Z. J. Zhang, X. L. Wu, S. H. Liu, X. D. Li, X. D. Sun and Y. Sakka, J. Am. Ceram. Soc., 2012, 95, 931.
4. C.-C. Chiang, M.-S. Tsai and M.-H. Hon, J. Electrochem. Soc., 2007, 154, J326.
5. J. K. Li, J.-G. Li, X. L. Wu, S. H. Liu, X. D. Li and X. D. Sun, Key Eng. Mater., 2013, 544, 245.
6. N. J. Cherepy, S. A. Payne, S. J. Asztalos, G. Hull, J. D. Kuntz, T. Niedermayr, S. Pimputkar, J. J. Roberts, R. D. Sanner, T. M. Tillotson, E. van Loef, C. M. Wilson, K. S. Shah, U. N. Roy, R. Hawrami, A. Burger, L. A. Boatner and W. S. Choong, IEEE Trans. Nucl. Sci., 2009, 56, 873.
7. K. Kamada, T. Endo and K. Tsutumi, Cryst. Growth Des., 2011, 11, 4484.
8. K. Kamada, T. Yanagida, J. Pejchal, M. Nikl, T. Endo, K. Tsutumi, Y. Fujimoto, A. Fukabori and A. Yoshikawa, J. Phys. D: Appl. Phys., 2011, 44, 505104.
9. J.-G. Li and Y. Sakka, Sci. Technol. Adv. Mater., 2015, 16, 014902.
10. J. K. Li, J.-G. Li, J. Li, S. H. Liu, X. D. Li, X. D. Sun and Y. Sakka, J. Solid State Chem., 2013, 206, 104.
11. J. K. Li, J.-G. Li, Z. J. Zhang, X. L. Wu, S. H. Liu, X. D. Li, X. D. Sun and Y. Sakka, Sci. Technol. Adv. Mater., 2012, 13, 035007.
12. J. K. Li, J.-G. Li, S. H. Liu, X. D. Li, X. D. Sun and Y. Sakka, Sci. Technol. Adv. Mater., 2013, 14, 054201.
13. J. K. Li, J.-G. Li, S. H. Liu, X. D. Li, X. D. Sun and Y. Sakka, J. Mater. Chem. C, 2013, 1, 7614.
14. J. Y. Choe, D. Ravichandran, S. M. Blomquist, D. C. Morton, K. W. Kirchner, M. H. Ervin and U. Lee, Appl. Phys. Lett., 2001, 78, 3800.
15. Y. Hakuta, K. Seino, H. Ura, T. Adschiri, H. Takizawa and K. Arai, J. Mater. Chem., 1999, 9, 267.
16. J.-G. Li, T. Ikegami, J.-H. Lee, T. Mori and Y. Yajima, J. Eur. Ceram. Soc., 2000, 20, 2395.
17. J.-G. Li, T. Ikegami, J.-H. Lee and T. Mori, J. Am. Ceram. Soc., 2000, 83, 961.
18. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
19. L. C. Thompson, Complexes, in Handbook on the Physics and Chemistry of Rare Earths, ed. K. A. Gschneidner Jr and L. Eyring, North-Holland Physics, Amsterdam, 1979.
20. Y. Zorenko, V. Gorbenko, T. Voznyak, T. Zorenko, B. Kuklinski, R. Turos-Matysyak and M. Grinberg, Opt. Spectrosc., 2009, 106, 365.
21. F. X. Xin, S. L. Zhao, S. Q. Xu, L. H. Huang, G. H. Jia, D. G. Deng and H. P. Wang, Opt. Mater., 2011, 34, 85.
22. F. Chen, M. Chen, C. Yang, J. Liu, N. Q. Luo, G. W. Yang, D. H. Chen and L. Li, Phys. Chem. Chem. Phys., 2015, 17, 1189.
23. G. H. Dieke and H. M. Crosswhite, Appl. Opt., 1963, 2, 675.
24. R. T. Wegh, A. Meijerink, R. J. Lamminmaki and J. Holsa, J. Lumin., 2000, 87–89, 1002.
25. P. S. Peijzel, A. Meijerink, R. T. Wegh, M. F. Reid and G. W. Burdick, J. Solid State Chem., 2005, 178, 448.
26. H. W. Song and J. W. Wang, J. Lumin., 2006, 118, 220.
27. K. Guo, M.-L. Huang, H.-H. Chen, X.-X. Yang and J.-T. Zhao, J. Non-Cryst. Solids, 2012, 358, 88.
28. S. Huang and L. Lou, Chin. J. Lumin., 1990, 11, 36.
29. Q. L. Dai, H. W. Song, M. Y. Wang, X. Bai, B. Dong, R. F. Qin, X. S. Qu and H. Zhang, J. Phys. Chem. C, 2008, 112, 19399.
30. L. G. van Uttert and L. F. Johnson, J. Chem. Phys., 1966, 44, 3514.
31. L. Ozawa, J. Electrochem. Soc., 1979, 126, 106.
32. P. Dorenbos, J. Alloys Compd., 2002, 341, 156.
33. N. C. George, K. A. Denault and R. Seshadri, Annu. Rev. Mater. Res., 2013, 43, 481.
34. M. Nikl, A. Vedda, M. Fasoli, I. Fontana, V. V. Laguta, E. Mihokova, J. Pejchal, J. Rosa and K. Nejezchleb, Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 76, 195121.
35. W. Chewpraditkul, L. Swiderski, M. Moszynski, T. Szczesniak, A. Syntfeld-Kazuch, C. Wanarak and P. Limsuman, IEEE Trans. Nucl. Sci., 2009, 56, 3800.
36. A. B. Munoz-Garcia, E. Artacho and L. Seijo, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 014105.
37. Y. Zorenko, T. Voznyak, V. V. Gorbenko, A. Doroshenko, A. Tolmachev, R. Yavetskiy, I. Petrusha and V. Turkevich, Opt. Mater., 2013, 35, 2049.
38. D. T. Haven, P. T. Dickens, H. M. Weber and K. G. Lynn, J. Appl. Phys., 2013, 114, 043102.
39. C. S. McCamy, Color Res. Appl., 1992, 17, 142.
40. R. Pang, C. Y. Li, L. L. Shi and Q. Su, J. Phys. Chem. Solids, 2009, 70, 303.
41. C. H. Huang and T. M. Chen, J. Phys. Chem. C, 2011, 115, 2349.
42. H. P. Christensen, D. R. Gabbe and H. P. Jenssen, Phys. Rev. B: Condens. Matter Mater. Phys., 1982, 25, 1467.
43. R. S. Meltzer, S. P. Feolov, B. Tissue and H. B. Yuan, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 60, 14012.
44. E. M. Campo, M. Pophristic, L. Hopkins and I. T. Ferguson, Appl. Opt., 2015, 54, 3613.
45. C. Feldman, Phys. Rev., 1960, 117, 455.
46. D. Kumar, K. G. Cho, Z. Chen, V. Craciun, P. H. Holloway and R. K. Singh, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 60, 13331.
47. G. Wakefield, E. Holland, P. J. Dobson and J. L. Hutchison, Adv. Mater., 2001, 13, 1557.

---