Rare-earth-doped glasses are known to be good candidate materials for devices such as optical fiber amplifiers, the solar cell conversion layers, solid state lasers, and so on.^[1–5] Among these glasses, the fluorozirconate (FZ) glass has the most efficient luminescence that originates from the following characteristics: high transmittance from the visible to middle infrared region, low phonon energy, and higher solubility of rare earth ions.^[6–8]

The substitution of chloride or bromide for fluorine in fluorozirconate glass around the rare earth ion changes the glass properties because the difference in ligand field symmetry between F–Cl and F–Br mixed-anion glasses. The presence of chloride or bromide helps to improve fluorescence emission efficiency because the phonon energies of chlorides and bromides are smaller than those of fluorides.^[9] The transformation of fluorochlorozirconate (FCZ) and fluorobromozirconate (FBZ) glasses to glass-ceramics can obviously improve the spectroscopic properties of the material and other properties.^[10]

Chloride or mixed halide glasses have a better transmission range due to the higher fundamental vibration frequency of the glass network.^[11] Soga et al.^[12] calculated the crystal field parameters of FCZ glass by a molecular dynamic simulation and point charge model of the crystal field; the results further showed that chlorine ions tended to coordinate to Eu³⁺ ion in zirconium rich glass and affected the luminescence properties of Eu³⁺ ions. Photoluminesence in Eu-doped fluorozirconate glass ceramic arises from Eu²⁺ ions in those BaBr₂ crystals, which has potential applications as x-ray storage phosphor.^[13]

Fluorescence spectra and Judd–Ofelt (JO) theory provide useful information for characterizing luminescence properties and coordination environments around active rare earth ion in glasses. In this paper, we chose the Eu³⁺ ion as the probe to discuss the luminescence behavior of fluorozirconate glass doped with Cl ions due to its different sensitivities to the ligand field between the magnetic dipole transition ⁵D₀ → ⁷F₁ and electric dipole transition ⁵D₀ → ⁷F₂. Due to the energy gap between ⁵D₀ and the next lower level being more than 1.2 × 10⁴ cm^{− 1}, luminescence properties of Eu³⁺ would hardly be affected by the phonon energy of the matrix. The influences on the structure and characteristic temperature parameters of FZ glass affected by Cl ions are analyzed by the differential thermal analysis (DTA) and the x-ray diffraction (XRD). We obtain the correlation among luminescence intensity, lifetime, and the ligand field symmetry through the fluorescence spectra, the fluorescence decay, and the Judd–Ofelt parameters.

The FZ/FCZ glass samples with the compositions of 53ZrF₄-10BaF₂-3.9LaF₃-3AlF₃-(10-x)NaF-xNaCl-10BaCl₂-0.1Eu₂O₃/53ZrF₄-20BaF₂-(4-y)LaF₃-3AlF₃-20NaF-yEu₂O₃/53ZrF₄-10BaF₂-4LaF₃-(4-y)AlF₃-10NaF-10NaCl-10BaCl₂-yEu₂O₃ (values in mol%) named FCZ-xNC (x = 0, 2, 4, 6, 8, 10), FZ-2yEu/FCZ-2yEu (y = 0.01, 0.1, 0.4, 0.8, 1.2, 1.6, 2.0, 2.4), were prepared by the meltquenching method. The Cl ion percentage concentration of total anions in FCZ-xNC glass is given in Table 1. The raw materials buried in the ammonium fluoride powders were thoroughly mixed in an alumina crucible and then heated at 800 °C for 30 min. The melting product was quickly transferred into a muffle furnace at a temperature of 200 °C under a reducing atmosphere. Followed by cooling to room temperature, the as-prepared transparent samples were presented. The samples were annealed at 260 °C in a reducing atmosphere for 2 h. Finally, cooling to room temperature, the annealed samples were cut and polished into dimensions of approximately 20 × 20 × 5 mm for measurement. The chemicals used in preparation of host glasses were of 99.95% purity in addition to Eu₂O₃ being of 99.99%.

Table 1.

Table 1.

Table 1. Cl ion percentage concentrations of total anions in FCZ-xNC glasses (x = 0, 2, 4, 6, 8, 10, 11). . FCZ-0NC FCZ-2NC FCZ-4NC FCZ-6NC FCZ-8NC FCZ-10NC FCZ-11NC Cl/(F+Cl)% 6.83 7.51 8.19 8.87 9.56 10.24 10.58

Table 1. Cl ion percentage concentrations of total anions in FCZ-xNC glasses (x = 0, 2, 4, 6, 8, 10, 11). .

Differential thermal analysis (DTA) was performed using a SHIMADZU DTG-60H with a 10-mg sample under an N₂ atmosphere at a heating rate of 10 °C/min, and the glass sample used for DTA measurements was powdered. The crystal structure of the samples was characterized by using a Rigaku 2500/PC x-ray diffractometer (XRD) with a Cu Ka radiation source (0.1542 nm) at 40 kV and 150 mA. Photoluminescence and time resolved spectrum measurements were performed by a Jobin Yvon FL3 fluorescence spectrophotometer with a 450-W xenon lamp and a flicker frequency 0.05 Hz–25 Hz flashing xenon lamp as an excitation source. All the measurements were carried out in ambient atmosphere.

The DTA curves of FZ/FCZ-6NC/FCZ-10NC/FCZ-11NC samples are given in Fig. 1, the T_g, T_x, T_p1 are glass transition temperature, initial crystallization temperature, and peak value of crystallization, respectively. The ΔT = T_x − T_g represents the devitrification tendency, the decreasing value of ΔT signifies the increasing devitrification trends and easier crystallization.^[14] These characteristic temperatures are listed in Table 2.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. Differential thermal analysis (DTA) curves of FCZ-xNC glasses (x = 0, 6, 10, 11).

Table 2.

Table 2.

Table 2. DTA parameters of FCZ-xNC glasses (x = 0, 6, 10, 11). . Tg/°C Tx/°C Tp1/°C ΔT/°C FZ 259 341 358 82 FCZ-6NC 258 343 360 85 FCZ-10NC 260 323 342 63 FCZ-11NC 277 378 404 101

Table 2. DTA parameters of FCZ-xNC glasses (x = 0, 6, 10, 11). .

The curve of characteristic temperature of the FCZ-6NC sample has little difference from that of FZ glass. Further increasing the concentration of Cl ions, the characteristic temperatures T_x and T_p1 of FCZ-10NC significantly decrease, and the ΔT also decreases. This result indicates that it is easier for crystallization to happen in the Cl-ions-rich sample. Additionally, the characteristic temperature of the FCZ-11NC sample further confirms this conclusion: the increasing of T_g temperature is due to the partial crystallization of the sample (as shown in Fig. 2) and the melting temperature of crystal is higher than that of the glass phase. The T_g is not changed when it is below a certain critical value, so we choose 260 °C as the annealing temperature, near the glass transition temperature.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. XRD patterns of FCZ-xNC glasses (x = 0, 6, 10, 11).

Figure 2 shows the XRD patterns of FZ/FCZ-6NC/FCZ-10NC/FCZ-11NC samples. Except FCZ-11NC, other samples all present two broad typical peaks at about 26° and 47°, indicating that FZ/FCZ-6NC/FCZ-10NC samples are in amorphous glassy states. Most studies have shown that fluorozirconate glass is mainly constitutive ZrF_n polyhedron,^[15] because the addition of Cl ions mainly replace F ions connected to the network modifier, and thus will not affect the FZ glass network structure.^[12] So the amorphous glassy state still remains in each of these samples. However, several peaks appear on the XRD pattern of the FCZ-11NC sample, which are assigned to mixed phases of BaZrF₆ and BaZr₂F₁₀ based on the standard XRD pattern. This phenomenon is due to cationic clusters that further cause the separation of the glass phase and the formation of the crystal phase.^[15] Meanwhile, these mixed phases lead to increasing the characteristic temperature of the FCZ-11NC sample in the DTA curve.

Figure 3 shows the photoluminescence (PL) spectra of FZ/FCZ-xNC (x = 0, 2, 4, 6, 8, and 10) doped with 0.2 mol Eu³⁺ and the variations of luminous intensity of FCZ-xNC with Cl ion concentration. Figure 3(a) shows 14 emission bands of Eu³⁺ in all the samples, the most intense emission bands are 613 nm (⁵D₀ → ⁷F₁) and 591 nm (⁵D₀ → ⁷F₂), the next is blue green light emission (⁵D_3,2,1,0 → ⁷F_4,3,2,1,0), radiative transition of ⁵D₀ → ⁷F_4,3 is relatively weak, luminous intensity of ⁵D₀ → ⁷F₀ is the weakest due to forbidding the selection rule. Additionally, the transitions bands of 417 nm–555 nm, which correspond to ⁵D_3,2,1 → ⁷F_4,3,2,1 are seldom observed in many glasses,^[16–18] which relate to the lower phonon energy in fluorozirconate glass (maximum phonon energy is 590 cm^{− 1}).^[19] With increasing the concentration of Cl ions, there is no obvious difference in emission location, but the emission intensity is obviously increased, FCZ-10NC is the highest one as shown in Fig. 3(b). The crystalline phase appears in the glass substrate of FCZ-11NC, thus there are no PL spectra of higher concentration samples. The magnetic dipole transition ⁵D₀ → ⁷F₁ peaked at 591 nm relates to Eu³⁺ centered at the inversion symmetry in the crystal structure, but the electric dipole transition ⁵D₀ → ⁷F₂ with a peak position of 613 nm relates to the Eu³⁺ at non-inversion symmetry. Figure 3(c) shows the curve of the R/O ratio, which is the ratio between intensities of ⁵D₀ → ⁷F₂ and ⁵D₀ → ⁷F₁ transitions versus Cl ion concentration, the gradually increasing R/O ratio indicates that more Eu³⁺ ions get into the non-inversion symmetry after replacing part of F ions by Cl ions.

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (a) PL spectra of FZ/FCZ-xNC glasses under 393-nm excitation, (b) luminous intensity analysis of FCZ-xNC glasses monitored at 613 nm, (c) R/O ratio of FCZ-xNC glasses, where x = 0, 2, 4, 6, 8, 10, 11.

In order to further study the influence of Cl ions on the local crystal field environment of Eu³⁺, we introduce the JO theory to calculate the parameters of the ⁵D₀ level of Eu³⁺ ions in FCZ-xNC glasses. The integral strength of emission spectrum and radiation transition probability with JO parameters have the following relationships:^[20]

Table 3.

Table 3.

Table 3. Values of JO parameters Ω2/Ω4, calculated lifetime τtal, experimental lifetime τexp, fitting lifetimes τ1 and τ2, fluorescence branching ratios A1 and A2, quantum efficiency η of Eu3+5D0 energy level for FCZ-xNC glasses (x = 0, 2, 4, 6, 8, 10). . FCZ-0NC FCZ-2NC FCZ-4NC FCZ-6NC FCZ-8NC FCZ-10NC Ω2/10−20 cm2 2.98 3.12 3.35 3.64 3.82 3.96 Ω4/10−20 cm2 2.46 2.36 2.09 2.15 2.42 2.52 τ1/ms 1.70 1.69 1.66 1.66 1.65 1.61 τ2/ms 5.03 5.06 5.13 5.14 5.17 5.23 A1/% 17.7 17.1 15.8 13.3 11.1 7.9 A2/% 82.3 82.9 84.2 86.7 88.9 92.1 τtal/ms 7.16 7.18 7.20 7.22 7.23 7.27 τexp/ms 4.80 4.84 4.93 4.97 5.03 5.14 η/% 67.0 67.4 68.5 68.8 69.6 70.7

Table 3. Values of JO parameters Ω2/Ω4, calculated lifetime τtal, experimental lifetime τexp, fitting lifetimes τ1 and τ2, fluorescence branching ratios A1 and A2, quantum efficiency η of Eu3+5D0 energy level for FCZ-xNC glasses (x = 0, 2, 4, 6, 8, 10). .

Figure 4 shows the normalized fluorescence intensity decay curves of FCZ-xNC glasses. All the curves increase with time obviously initially. The risetime can be explained by the excited-state absorption of higher levels (⁵D₁) or energy transfer from neighboring Eu³⁺ ions.^[19] All curves fit to double-exponential decay, the average lifetime of Eu³⁺ ions can be expressed as

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. Normolized fluorescence decay curves of FCZ-xNC glasses at room-temperature of the 5D0 → 7F2 transition excited at 393 nm, monitored at 613 nm. Inset: double-exponential fitted lifetime of FCZ-xNC glasses, where x = 0, 2, 4, 6, 8, 10, 11.

Figure 5 shows the luminous intensity point diagram of different concentrations of Eu³⁺ in FZ-2yEu and FCZ-2yEu glass samples. Under the same concentration of Eu³⁺, luminous intensities of all the glasses containing the Cl ion increase. We find that the Cl ion causes no change of the best Eu³⁺ doping concentration (peaked at 3.2 mol) in FZ glass, which demonstrates that the introduction of Cl ion does not affect the Eu³⁺ solubility in the glass matrix.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. Variations of luminous intensity of 5D0 → 7F2 transition in FZ and FCZ-10NC glasses with the concentration of Eu3+ ions excited at 393 nm.

Figures 6(a) and 6(b) show the time-dependent normalized fluorescence decay curves of FZ-2yEu and FCZ-2yEu glasses for different concentrations of Eu³⁺ ions, respectively. Fitting results indicate that fluorescence decays of FCZ glass are slightly higher than that of FZ glass under the same concentration of Eu³⁺ ions as shown in Fig. 6(c). With increasing the concentration of Eu³⁺ ions, the average lifetime τ_exp (hollow) first increases then decreases. This change of lifetime is consistent with that of luminous intensity. The lifetime decreases when the Eu³⁺ concentration is over 3.2 mol due to the self-quenching. The best doping concentration corresponds to the longest lifetime. The introduction of Cl ions does not affect the variation trend of lifetime. The lifetime τ_cal (solid) calculated from JO theory is given at the same time, and the theoretical simulation is in accordance with the experimental result. Figure 6(d) gives the estimated quantum efficiency, showing that the quantum efficiency is increased by 3∼4% in all FCZ-2yEu glasses due to the introduction of Cl ions.

Fig. 6.

Figure Option ViewDownloadNew Window

Fig. 6. (a) normolized fluorescence decay curves of FZ-2yEu (y = 0.02–4.8 mol) glasses, (b) normolized fluorescence decay curves of FCZ-2yEu (y = 0.02–4.8 mol) glasses excited at 393 nm, monitored at 613 nm, (c) calculated lifetime τtal (solid), and experimental lifetime τexp (empty) values for FZ-2yEu and FCZ-2yEu glasses, (d) quantum efficiency η values of FZ-2yEu and FCZ-2yEu glasses respectively.

A series of Eu³⁺ ions doped fluorozirconate glasses containing chloride ions are prepared. The introduction of Cl ions reduces the glass crystallization temperature, and increases the devitrification trend of fluorozirconate glass, leading to the crystalline phase forming more easily. Cationic clusters lead to the separation of the glass phase and the generation of the BaZrF₆/BaZr₂F₁₀ phase, when the Cl ion concentration exceeds 10.24%. The luminous intensity of Eu³⁺ ions is dramatically enhanced with the increase of Cl ion concentration, and the quantum efficiency is increased from 67.0% to 70.7% at the same time. The doped Cl ions can reduce the local symmetry of Eu³⁺ ions, indicated by analyzing the R/O ratio and JO parameters. Fluorescence decay analysis reveals that there are two kinds of sites for Eu³⁺ ions in a glass network, corresponding to short-lifetime and long-lifetime, respectively. The average lifetime increases with increasing the concentration of Cl ions. Moreover, the proportion of Eu³⁺ with a long-lifetime is improved, while that with the short-lifetime is opposite. Additionally, the substitution of the Cl ion for the F ion does not affect the solubility of rare earth Eu³⁺ ion nor the variation trend of fluorescence decay under different Eu³⁺ concentrations, and the best doping concentration corresponds to the longest fluorescence decay. In a word, the introduction of the Cl ions can strengthen the luminous intensity of glass, increase the average lifetime and improve quantum efficiency by about 3%–4%.