1. Introduction

In recent years, mixed-anion compounds have attracted significant attention owing to their various functionalities, which originate from their unique structural and electronic properties [1–4]. These properties are induced by the incorporation of multiple anions with different ionic radii, electronegativities, oxidation states, and polarizabilities. One of the most important aspects of studies based on mixed-anion compounds is to control and understand the arrangement of anions that determine their electronic structures, for example, the indirect bandgap characteristics of Sr₂ScO₃Cl originating from ScO₅ pyramids [5] and two-dimensional quantum antiferromagnetism in Sr₂CuO₂Cl₂ with trans-configuration of Cl ions in the CuO₄Cl₂ octahedra [6].

Ruddlesden–Popper (RP)-type layered perovskite with a general chemical formula A_n+1B_nX_3n+1 (where A = electropositive cation, B = transition metal, X = anion, and n = perovskite block number) is an interesting stage for studying the relationship between the anion order and physical and chemical properties [7–9]. The ideal RP-type structure contained more than one different anion site. For example, for the n = 1 phase, the apical and equatorial anion sites of the octahedra are surrounded by AB₅ and A₂B₄ polyhedra, respectively, which differ from six equivalent sites with A₄B₂ coordination in the ideal cubic perovskite system ABX₃. Therefore, if there are large differences in these anions, this phase can form ordered arrangements.

In our previous study, we reported the synthesis, crystal structure, and optical properties of a new RP-type layered perovskite oxychloride, Ba₃Y₂O₅Cl₂ [4]. The crystal structure is shown in Fig. 1. The Y atom in this compound shifted toward the apical oxygen site owing to the asymmetric coordination of O and Cl. The direct bandgap of the compound is ∼5.4 eV. In addition, we investigated the luminescence properties with Eu doping. As Y³⁺ ions with a distorted octahedron have an asymmetric coordination environment, both magnetic and electric dipole transitions were observed by doping Eu³⁺ to Y³⁺ sites, which gives characteristic orange luminescence. This compound also exhibited a high internal quantum yield (IQY) (>70%). Considering the characteristic luminescence and high IQY, this compound is a potential host lattice for luminescent materials.

 

Fig. 1. Crystal structure of Ba₃Y₂O₅Cl₂. Gray, blue, red, and green balls represent Ba, Y, O and Cl ions, respectively. The blue pyramids show polyhedral of YO₅.

Download Full Size | PDF

It is well known that the Tb³⁺ ion shows green emission when doped into wide-gap materials [10,11]. In addition, it emits some sharp lines in a wide visible range, and the most intense line at approximately 545 nm is frequently utilized in green lighting. Luminescence characteristics, such as peak positions, peak shapes, and decay time, are all influenced by the crystal field and site symmetry. These characteristics can also be observed when perovskites and/or layered perovskites are used as host lattices [12,13]. For Tb³⁺-doped LaScO₃ perovskite, Tb³⁺ at B sites shows green luminescence; in contrast, Tb³⁺ at A sites shows blue-violet and green luminescence. The decay time of luminescence from the B sites was systematically longer than that from the A sites [14].

For phosphors with a perovskite-related matrix, it is common to locate the luminescence center to the A site, whereas doping at the B site is relatively unusual [15]. In order to accommodate lanthanide ions at B sites, not only B ions but also A ions are required to be large [16], similar to the case for Ba₃Y₂O₅Cl₂. In addition, owing to its special structural and electronic properties, it is interesting to investigate the luminescence of Tb³⁺-doped Ba₃Y₂O₅Cl₂.

In this study, we synthesized Ba₃Y₂O₅Cl₂ with different Tb³⁺ doping concentrations. The luminescence properties were studied and compared with those of the Y₂O₃: Tb³⁺ phosphors. We also proposed the luminescence mechanisms and concentration quenching.

2. Experimental procedure

2.1 Synthesis

Polycrystalline samples of Ba₃Y₂O₅Cl₂ were synthesized by a conventional solid-state reaction with stoichiometric amounts of BaCO₃ (3N), BaCl₂ (3N), and Y₂O₃ (3N), according to our previous report [4]. Powder mixtures were pelletized and heated one or two times at 1173 K in air for 24 h with intermittent grindings.

Polycrystalline samples of Ba(Y_1-xTb_x)₂O₅Cl₂ (Ba₃Y₂O₅Cl₂:Tb³⁺, x = 0-0.10) were synthesized by the same method with stoichiometric amounts of BaCO₃ (3N), BaCl₂ (3N), Y₂O₃ (3N) and the corresponding amount of Tb₄O₇ (3N) and heating at 1248 K under 5% H₂/Ar for 12 h. Y₂O₃ and Tb₄O₇ were used as starting materials and fired at 1773 K in air for 24 h, to prepare Y₂O₃:Tb³⁺ powder for comparison.

2.2 X-ray diffraction

The sintered samples were measured by powder X-ray diffraction (PXRD) with an Ultima-IV diffractometer (Rigaku). The data were collected in the range 5° ≤ 2θ ≤ 80° with a step size of 0.02° using Cu Kα radiation. The lattice parameters were refined using PXRD patterns with a Si internal standard.

2.3 Optical properties

Photoluminescence spectra were collected using a spectrofluorometer (JASCO, FP-8500). All the optical measurements were performed at room temperature.

3. Results and discussion

Ba₃Y₂O₅Cl₂:Tb³⁺ samples were successfully prepared by solid-state synthesis, as reported previously. Figure 1 shows the crystal structure of Ba₃Y₂O₅Cl₂. Figure 2(a) shows the XRD patterns of Tb-doped Ba₃Y₂O₅Cl₂ samples. Almost a single phase was formed, except for a small amount of Y₂O₃ impurity. Considering that there is no Tb₄O₇ impurity, we can speculate that Tb ions were successfully doped into the matrix. The lattice parameters for different Tb concentrations are plotted in the inset of Fig. 2(b). The lattice parameters of undoped sample were a = 4.3971(8) Å and c = 24.848(5) Å [4]. Since ionic radius of Y³⁺ (0.9 Å, coordination number (C.N.) 6) and Tb³⁺ (0.923 Å, C.N. 6) are similar [17], Tb³⁺ doping does not affect the lattice constants too much, except for a slight increase in the a-axis. For the c-axis, the lattice constants show large disturbances, especially for the steep drop of 10%- doped sample. This is consistent with the crystal structure of this compound. Y-Cl is weakly bonded compared to Y-O; therefore, with increasing Tb concentration, doping expands only the a-axis but not the c-axis [4,12]. The unit cell volumes (V_cell) calculated from the lattice parameters with different Tb concentrations are plotted in Fig. 2(b). We can observe that the general trend increases with increasing Tb concentration. The increased cell volume is consistent with related rare earth ions doing at B site rather than A site [12–14], which is in accord with the specific photoluminescent property originates from the B site ions that we will discuss later.

 

Fig. 2. (a) XRD patterns and (b) unit cell volumes of Ba₃Y₂O₅Cl₂:xTb³⁺ (x = 0, 1, 2,3, 4, 6, 8, and 10%). Inset is lattice parameters of Ba₃Y₂O₅Cl₂: xTb³⁺ (x = 0-10%).

Download Full Size | PDF

Figure 3 shows the excitation and emission spectra of Ba₃Y₂O₅Cl₂:1% Tb³⁺ and Y₂O₃:1% Tb³⁺ at room temperature. The excitation spectrum was recorded by monitoring with bright green emission at 544 nm owing to the 4f⁸ −4f⁷5d¹ transition of the Tb³⁺ ions. Both of the samples show excitation peak characteristics of the 4f−5d transition of Tb³⁺. We also observed several weak peaks in the range of 300–500 nm, which were caused by the spin and orbital forbidden intra-4f transitions (see inset in Fig. 3) [18]. For the Ba₃Y₂O₅Cl₂:Tb³⁺ sample, the broad peak at 238 nm is the host lattice excitation peak, which is consistent with the optical band gap of the Ba₃Y₂O₅Cl₂ host. Another broad peak at 290 nm can be ascribed to the 4f⁸ -4f⁷5d¹ transitions of Tb³⁺ ions. In the case of the 4f→5d transitions of Y₂O₃:Tb³⁺, there are two peaks at ∼278 and 304 nm, whereas in Ba₃Y₂O₅Cl₂:Tb³⁺, there is only a peak at ∼290 nm. The shape of 4f→5d band varied, indicating different host lattice structures. In Y₂O₃, there are two different sites of Y³⁺, whereas in Ba₃Y₂O₅Cl₂, there is only one site [4].

 

Fig. 3. Photoluminescence of Ba₃Y₂O₅Cl₂:1% Tb³⁺(up) and Y₂O₃:1% Tb³⁺ (down) measured at room temperature. The dashed and solid lines show the excitation spectrum and emission line, respectively (inset: expanded spectrum of excitation spectrum of Ba₃Y₂O₅Cl₂:1% Tb³⁺).

Download Full Size | PDF

In the emission spectrum, when Ba₃Y₂O₅Cl₂:Tb³⁺ and Y₂O₃:Tb³⁺ were excited at wavelengths of 290 and 304 nm, respectively, photoluminescence from Tb³⁺ 4f−4f transitions were observed in both samples. The spectra show emission lines peaking at 487-498, 539-560, 585-597, and 616-635 nm due to transitions of ⁵D₄→⁷F₆, ⁵D₄→⁷F₅, ⁵D₄→⁷F₄, and ⁵D₄→⁷F₃ levels from Tb³⁺, respectively. The most intense emission at 544 nm stems from the ⁵D₄→⁷F₅ transition, which is a green emission. These results are consistent with the results for Tb³⁺ ions in other hosts, such as Y₃Al₅O₁₂ and Lu₂O₃ [19–21]. The ⁵D₃→⁷F_J emissions were not observed in our experiments. It is noteworthy that the highest peak at 544 nm splits into two peaks for both the samples. Ba₃Y₂O₅Cl₂:Tb³⁺ splits to 544 and 554 nm for Y₂O₃:Tb³⁺ at 544 nm and 552 nm, respectively. Further, we observe that the intensity ratio of split peaks is different from that of Y₂O₃:Tb³⁺. This type of peak split is due to the effect of crystal field energy. The split width for Ba₃Y₂O₅Cl₂:Tb³⁺ was 10 nm, while that for Y₂O₃:Tb³⁺ was 8 nm. This larger split width suggests that the crystal field energy of Ba₃Y₂O₅Cl₂:Tb³⁺ is larger than that of Y₂O₃:Tb³⁺ [11]. For Ba₃Y₂O₅Cl₂, the average bond length of Y-O is 2.2038 Å, whereas that for Y₂O₃ is 2.2842 Å, which is much larger than that for Ba₃Y₂O₅Cl₂ [4]. The closer the distance of Y and O ions, the stronger the repulsion of the O ions and the ${d_{{x^2} - {y^2}}}$ $\textrm{and}\; {d_{{z^2}}}\; $ orbitals of Y ions, causing the larger crystal field energy.

Figure 4 shows the PLE and PL spectra with different Tb concentrations. With an increase in Tb concentration, the intensity increases gradually, reaching a maximum value when the Tb³⁺ concentration is up to 8%. When the doping concentration is beyond 8%, the intensity decreases significantly. For increasing dopant concentrations, higher light outputs are obtained due to the larger number of luminescent centers. Up to a critical point, the intensity decreases. In addition, the spectral shape and locations of the peaks do not vary with the doping concentrations of Tb³⁺ ions for Ba₃Y₂O₅Cl₂:Tb³⁺. From the normalized PLE spectra, as shown in the inset of Fig. 4(a), we can observe a slight red shift of the f-d excitation band for increasing Tb concentrations. The Tb³⁺ has [Xe]4f⁸ electronic configuration, the excitation of Tb³⁺ originates from 4f⁸ to 4f⁷5d¹ transition. Because of the shielding by the outer 5s and 5p orbitals, the 4f electron is insensitive to the crystal field. However, the 5d electron interacts strongly with the crystal lattice, resulting in strong phonon coupling and large crystal field effects on the excited states [22]. In addition, we observe that the excitation bands of f-d excitation become broader with the increment of Tb³⁺ concentration. It’s known that the crystal field splitting energy of 4d element is larger than that of 3d element. Therefore, this broader bands can be assigned to the larger splitting of energy levels as a result of the larger crystal field energy caused by Tb³⁺ doping.

 

Fig. 4. The room temperature (a) PLE spectra and (b) PL spectra of Ba₃Y₂O₅Cl₂:xTb³⁺ (x = 1, 2, 3, 4, 6, 8, and 10%) phosphors (λ_ex= 290 nm). Inset of (a) shows the normalized PLE spectra, inset of (b) shows the expanded spectra of the emission spectra.

Download Full Size | PDF

Figure 5(a) shows the photograph of Ba₃Y₂O₅Cl₂:xTb³⁺ under excitation by UV lamp. Figure 5(b) shows the normalized intensities of the strongest emission peak (544 nm) with different Tb concentrations. Compared with simple perovskite host materials, this quenching concentration is very high 12; this can be attributed to the layered structure of Ba₃Y₂O₅Cl₂. Tb³⁺ occupies the center sites of the YO₅ polyhedron, and Ba-Cl layers can effectively isolate each Tb³⁺. The relatively large spatial distance between Tb³⁺ ions in the layers results in a reduction in nonradiative energy between sites, which serves to suppress concentration quenching effect [23]. This can be further demonstrated by Density Functional Theory (DFT) calculation result in our previous study 4. When x = 8%, the luminescence center is just the right saturation, as the emission and excitation intensity are at their strongest values, which are higher than those at other concentrations. When x >8%, Tb³⁺ ions position close to each other, and the probability of interaction between adjacent two Tb³⁺ ions increases; this, in turn, increases the new energy loss, thereby resulting in dislocation and cross-relaxation phenomenon significantly weakening the luminescence properties of phosphors [24]. With layered structure separating active ions, quenching can be hindered [25]. Therefore, the quenching concentrations of this compound are relatively higher.

 

Fig. 5. (a) Ba₃Y₂O₅Cl₂:xTb³⁺ under excitation by UV lamp (l_em. 250 nm); and (b) relation between PL intensity and Tb³⁺ concentration.

Download Full Size | PDF

The emission intensity decreased because of the energy transfer between the Tb³⁺ ions. In general, there are two main kinds of energy transfer: one is exchange interaction, wherein the typical critical distance is no more than 5 Å, and the other is multipolar interaction, wherein the critical distance is longer than 10 Å [19,20]. The distance between adjacent ions when concentration quenching occurs is called the critical distance (Rc). Rc can be calculated using the following formula:

 

Fig. 6. Relationship of log (I/x) vs. log (x)for Ba₃Y₂O₅Cl₂:Tb³⁺ phosphor.

Download Full Size | PDF

The luminescence lifetime at the ⁵D₄ → ⁷F₅ transition in Ba₃Y₂O₅Cl₂:8%Tb³⁺ is shown in Fig. 7(a). The decay curve I(t) was fitted using the following equation:

 

Fig. 7. (a) Luminescence decay of Ba₃Y₂O₅Cl₂:8% Tb³⁺; (b) lifetimes of Ba₃Y₂O₅Cl₂:xTb³⁺ (x = 0.01–0.1). The excitation and monitoring wavelength are 290 and 544 nm, respectively.

Download Full Size | PDF

Figure 8 shows the proposed luminescence mechanisms in Ba₃Y₂O₅Cl₂:Tb³⁺. Tb³⁺ can be excited through 4f⁸ → 4f⁷5d¹ transitions. Empty Tb 5d orbitals are expanded more than 4f orbitals and easily influenced by the crystal field. Due to the crystal field energy of Ba₃Y₂O₅Cl₂, the empty Tb 5d orbitals have larger splitting. The bottom of Tb³⁺ 5d orbitals shift near ⁵D₃ regimen, whereas they shift below the ⁵D₃ excited states by relaxation just after excitation, that is ⁵D₃ energy levels in Ba₃Y₂O₅Cl₂:Tb³⁺ lie above the band gap or within the conduction band. Therefore, it can give rise only to green luminescence originated from ⁵D₄ to ⁷F_J, and no energy transfer occurs from the 5d orbitals to ⁵D₃ excited states, thus ⁵D₃ to ⁷F_J emission cannot be observed.

 

Fig. 8. Schematic luminescence mechanisms in Ba₃Y₂O₅Cl₂:Tb³⁺ on 4f–5d excitation.

Download Full Size | PDF

4. Conclusion

Herein, we successfully synthesized Tb-doped Ba₃Y₂O₅Cl₂ via a solid-state reaction. Green emission from the Tb³⁺ 4f−4f transitions was observed. Peak splitting of green ⁵D₄luminescence was observed. The width of split peaks of Ba₃Y₂O₅Cl₂ is larger than that of Y₂O₃ owing to the stronger crystal field energy of Ba₃Y₂O₅Cl₂. In addition, the photoluminescence at different concentrations was investigated. Concentration quenching at 8% was found, which is much higher than that in simple perovskite host materials. This concentration quenching is due to the dipole-dipole interaction between neighboring Tb³⁺ ions. Further, luminescence mechanisms have been proposed. Layered perovskite structure, Ba₃Y₂O₅Cl₂, can be host lattice for different luminescence centers, and also hinder luminescence quenching, therefore, it is promising to be used in optical devices.