1. Introduction

Ferroelectric titanates such as SrTiO₃ and BaTiO₃ (BTO) with typical perovskite ABO₃ structure have been extensively studied due to their excellent dielectric, ferroelectric, and electro-optic properties [1–4]. Er³⁺ as an active ion has drawn broad attention to meet the request used for upconversion (UC) phosphors, planar waveguide and structural probe [5–8]. BTO doped with Er³⁺ ion has been extensively investigated in ferroelectricity, phase transitions and luminescence. It is noted that the radius of the Er³⁺ ion is intermediate between that of the Ba²⁺ ion and Ti⁴⁺ ion. Er³⁺ can occupy either A- or B-site depending on Ba/Ti ratio [9–12]. Dopant site location was found to play an important role in electrical properties of Er³⁺-doped BTO [11–13]. It is known that crystal field caused by structure symmetry of the host materials would contribute to different perturbation in terms of the Er³⁺ inner shell transitions. Therefore, UC photoluminescence (PL) efficiency should be dependent on the excited-state dynamics of the Er³⁺ ions and their interactions with the host matrix [14,15]. The interactions are strongly influenced by the host and the dopant. Unfortunately, there is very limited knowledge about the influence of site substitution of Er³⁺-doped BTO on its PL properties. Furthermore, BTO is a typical ferroelectric material that undergoes successive phase transitions under different temperatures. UC mechanism of substituting sites corresponding to different symmetry structures has not been investigated yet. In this work, BTO doped with different Er³⁺ substituting positions and concentrations were prepared. X-ray diffraction (XRD) and Raman spectra were used to reveal the lattice distortion with Er³⁺ doped. The PL spectra of Er³⁺-doped BTO and the UC mechanism related to the structure and cross-relaxation (CR) were studied. The PL spectra as a function of measuring temperature were investigated in order to understand the relationship between phase transition and luminescent behavior.

2. Experiment

Er³⁺-doped BTO powders with different site substitutions and concentrations were prepared by solid-state reaction method. Reagent grade BaCO₃, TiO₂, and Er₂O₃ powders were used as raw materials. The charge compensation could be compensated by barium and oxygen vacancy for the A- and B-site substitutions, respectively. Here the samples for A- and B-site substitutions are refered to Ba_1-3x/2Er_xTiO₃ (BTO: A) and BaTi_1-xEr_xO_3-x/2 (BTO: B), respectively. Based on the above formulas, the starting powders with designed stoichiometric quantities were ball milled for 24 h, then dried and calcinated at 1100 °C for 8 h in air to generate the Er³⁺-doped BTO powders.

The crystal structure of the samples was examined by a Bruker D8 Advance X-ray diffractometer (XRD). Raman spectra were measured using a Horiba Jobin Yvon HR800 Raman spectrometer with a 488 nm laser excitation source. The PL spectra were recorded using an Edinburgh FLSP920 spectrophotometer under the excitation of a 980 nm laser diode. The fluorescence lifetimes for Er³⁺ were recorded at 552 nm with a μF900H micro-second flashlamp (λ_ex = 488 nm) as the excitation sources. The temperature-dependent PL spectra were carried out in the temperature range of 15-300 K.

3. Results and discussion

Figure 1 shows the XRD patterns of Er³⁺-doped BTO with different site substitutions and concentrations. The characteristic diffraction peaks of tetragonal BTO phase without secondary impurity phases were observed. The result implied that Er³⁺ ions were doped efficiently into BTO host. A comparison of these XRD patterns with various Er³⁺ doping shows the broaden diffraction peaks for samples with higher Er³⁺ doping concentrations as shown in Fig. 1. It suggests that the doping with higher Er³⁺ concentrations may lead to the worse crystalline of doped BTO. In addition, compared with pure BTO, there are minor shifts of the diffraction peak (111) for the Er³⁺ doped samples as seen in the inset of Fig. 1. It means that the lattice constant of Er³⁺-doped BTO shrinks or expands in the case of Er³⁺ occupying A- or B-site, respectively [12,19].

 

Fig. 1 XRD patterns of the pure BTO and Er³⁺-doped BTO with different site substitutions and concentrations: (a) pure, (b) x = 3 mol% with A-site substitution, (c) x = 3 mol% with B-site substitution and (d) x = 5 mol% with B-site substitution. The inset shows the enlarged (111) peak.

Download Full Size | PDF

Figure 2 gives the Raman spectra of Er³⁺-doped BTO and pure BTO at room temperature. BTO exhibits tetragonal structure belonging to the space group C_4v symmetry. All of the features observed in the tetragonal phase have been reported in the literature [16]. The peak observed at 305 cm⁻¹ corresponds to the E(TO₂) phonon mode of tetragonal BTO. The A₁(TO₁), A₁(TO₂), A₁(TO₃) and A₁(LO₃) modes were observed at about 180, 270, 516 and 720 cm⁻¹, respectively. Raman spectra obtained from Er³⁺-doped BTO did not show any remarkable wavelength shift. It can also be seen that all Raman modes become weaker and broader with an increase in Er³⁺ concentration. It indicates the higher Er³⁺ concentration results in the worse crystallinity, which is consistent with the XRD results.

 

Fig. 2 Raman spectra of the pure BTO and Er³⁺-doped BTO with different site substitutions and concentrations.

Download Full Size | PDF

Figure 3 shows the PL spectra of BTO doped with different Er³⁺ ion concentrations and site substitutions measured at room temperature. The typical UC emission consists of two strong green bands located at 523 and 552 nm corresponding to ²H_11/2/⁴S_3/2→⁴I_15/2 transitions, and a weak red emission band at 656 nm ascribed to ⁴F_9/2→⁴I_15/2 transition of the Er³⁺ ion, respectively. Obvious Stark-splitting can be observed in all samples. Note that with an increase in Er³⁺ doping concentration (0.5-5 mol% for BTO: B), the emission intensity reaches a maximum value at 1 mol%, and then decreases with an increase in doping concentration. There are two reasons responsible for this observation. Structural analysis based on XRD and Raman spectra has revealed that worse crystalline occurs in BTO: Er with increasing Er³⁺ content. Worse crystalline phase corresponds to higher defect density of the materials. Both subband gap defect levels and impurity atoms in the grain boundaries could participate in the relaxation process and change the probability of radiative recombination [17,18]. On the other hand, an increased dopant concentration could also enhance CR process with remarkable decrease UC emission as shown in Fig. 4 . Under the excitation of 980 nm, through excited state absorption (ESA) or energy transfer (ET) process, Er³⁺ ion can populates the ⁴F_7/2 level. Subsequently, the Er³⁺ ion then relaxes nonradiatively to the ²H_11/2 and ⁴S_3/2 levels by multiphonon relaxation, from which the green ²H_11/2/⁴S_3/2→⁴I_15/2 emissions occur. As shown in Fig. 3, red UC emissions from the ⁴F_9/2 state are weak, since the ⁴F_9/2 state has a relatively large energy separation below the ⁴S_3/2 state, compared with the phonon energy of the BTO lattice. Moreover, we calculated the intensity of green-to-red ratio (I_523nm/I_656nm) for the BTO: B sample as shown in the inset of Fig. 3. It can be seen that the ratio (I_523nm/I_656nm) is 23.4, 24.2, 5.7, and 1.8 for 0.5, 1, 3, and 5 mol% B-site Er³⁺-doped BTO, respectively. The result can be understood based on the CR process shown in Fig. 4. Higher Er³⁺ concentration contributing to a shorter distance between ions results in an enhanced ET probability of the CR process, i.e. ²H_11/2 + ⁴I_15/2→⁴I_9/2 + ⁴I_13/2. Hence, the Er³⁺ ions at ⁴I_13/2 state through ESA populate the ⁴F_9/2 state. This enhanced population of the ⁴F_9/2 state primarily enhances the intensity of the red emission.

 

Fig. 3 The PL spectra of BTO: A with x = 3 mol% (red) and BTO: B with x = 0.5 mol% (magenta), 1 mol% (cyan), 3 mol% (blue) and 5 mol% (black). The inset shows the green-to-red ratio (I_523nm/I_656nm) vs dopant concentration.

Download Full Size | PDF

 

Fig. 4 Schematic diagram for the UC mechanism of Er³⁺ ions under 980 nm excitation.

Download Full Size | PDF

The PL spectra of BTO with 3 mol% Er³⁺ ion doped for A- and B-site substitutions are also presented in Fig. 3. The PL intensity differs for different site substitutions, but there is no obvious shift of the emission peaks. These results indicate that the crystal field surrounding the erbium ions can change the relative radiative probability of the transitions from the excited state to the eight Stark components of the ⁴I_15/2 ground state. The PL intensity of A-site doped sample is much stronger than that of B-site doped BTO. Compared with that of A-site doped sample, the decrease in PL intensity for B-site substitution Er³⁺-doped BTO may be affected by oxygen vacancies when Er³⁺ as an acceptor type impurity replaces Ti site, leading to an oxygen-deficient nonstoichiometry [13,19]. Compared the fluorescence lifetimes for two samples at 552 nm corresponding to ⁴S_3/2→⁴I_15/2 transition, the double exponential fitting results for A- and B-site doped samples are 132 and 52 ms, respectively. The decrease of lifetime of B-site substituted sample could be ascribed to the introduction of oxygen vacancies. We speculate that these oxygen vacancies would induce mew energy levels, which promote nonradiative energy transfer.

It is well-known that BTO undergoes successive phase transitions. PL evolution of the samples has been evaluated as a function of temperatures between 15 and 300 K as shown in Fig. 5 . It shows that the PL response of Er³⁺ ions to the structural changes occurring during phase transition is related to their site location in the BTO host. Figure 5(a) shows the temperature dependence of the PL intensity for the case of A-site substitution. It is known that pure BTO crystal undergoes ferroelectric phase transitions at 278 and 183 K. Rare-earth doping has slight influence on the phase transition temperature of BTO [20]. As presented in Fig. 5(b), the abnormal jump in PL intensity across the vicinity of 150 K and 250 K may be related to the crystal field changes around Er³⁺ ions induced by the phase transition. For the B-site substitution, Fig. 5(c) illustrates that the UC emission from ²H_11/2 is strongly suppressed, while the emission from ⁴S_3/2 increases remarkably. The ²H_11/2 state is thermally quenched to ⁴S_3/2 state and subsequently contributed to the enhanced population of ⁴S_3/2 state with decreasing temperature. Apart from the thermal quenching, the phenomenon is also relative to the crystal field. The crystal field formed by the octahedral oxygen ions with a lower symmetry than O_h is more suitable for the Er intra-4f transitions [21,22]. Therefore, Er³⁺ located B-site with lower symmetry due to phase transition facilitates the UC emissions. An increase in red emission ⁴F_9/2→⁴I_15/2 with lower symmetry occurs as evidenced in Fig. 5(d). The characteristics imply that the emission from substituted Er³⁺ ions can be used as a structural probe for the BTO.

 

Fig. 5 (a) and (c) show the PL spectra of Er³⁺-doped BTO (x = 3 mol%) with different substitutions at different temperatures 15 K (black), 50 K (red), 100 K (blue), 150 K (cyan), 183 K (magenta), 250 K (yellow) and 300 K (navy). (b) and (d) show the corresponding integrated intensity of three emissions at different temperatures. (a) and (b) for A-site substitution, (c) and (d) for B-site substitution.

Download Full Size | PDF

4. Conclusions

In conclusion, PL quenching of Er³⁺-doped BTO is due to the worse crystalline and cross-relaxation with an increase in Er³⁺ doping concentration. Phase transition results in the increase of PL spectra of A-site doped BTO across the phase transition temperature. For the case of B-site substitution, ⁴S_3/2→⁴I_15/2 emissions are strongly enhanced due to thermal quenching and lower symmetry induced by phase transition.