Tunable emission, energy transfer, and thermal stability of Ce³⁺-doped and Ce³⁺/Tb³⁺ Co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors

Abstract

Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors were synthesized successfully via a traditional high-temperature solid-state reaction. The crystal structure, photoluminescent properties including temperature-dependent luminescence, and energy transfer of the as-prepared phosphors were investigated. The as-prepared Ca₉Sr(PO₄)₆Cl₂:Ce³⁺ phosphors exhibit a broad excitation band ranging from 220 to 385 nm and blue light-emitting band centered at 431 nm, which originate from the 4f–5d transitions of Ce³⁺ ion. The luminescent intensities of Tb³⁺ ions were dramatically enhanced by the introduction of Ce³⁺ in the Ca₉Sr(PO₄)₆Cl₂:Tb³⁺ phosphors because of the efficient energy transfer from Ce³⁺ to Tb³⁺ ions, generating tunable blue-green emission colors. The mechanism of energy transfer between Ce³⁺ and Tb³⁺ ions was demonstrated to be an electric dipole–quadrupole interaction. Moreover, the energy transfer efficiency was evaluated up to 75 % based on the analysis of the emission spectra. The temperature-dependent photoluminescence indicates that the as-prepared Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors have excellent thermal stability. Our results suggest that the Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors have potential application for n-UV pumped WLEDs.

Introduction

White light-emitting diodes (WLEDs) are considered as an ideal candidate for the replacement of conventional lighting resource because of their remarkable advantages such as long operation lifetime, high luminous efficiency, energy saving, and eco-friendliness [1–3]. Typically, the commercial WLEDs were assembled by a combination of a blue-emitting InGaN chip and a yellow-emitting Y₃Al₅O₁₂:Ce³⁺ (YAG:Ce³⁺) phosphor. However, this type of device suffers from inevitable drawbacks including a low color rendering index (CRI < 75) and high correlated color temperature (CCT ≥ 4500 K) due to the lack of red component in the visible region [5–7]. Nowadays, in view of these problems, WLEDs fabricated by assembling ultraviolet (UV) or near-ultraviolet (n-UV) LED chips and tricolor (red, green, and blue) phosphors are expected to dominate the solid-state lighting (SSL) market in the future because the white light outputting from this type of WLEDs has controlled color temperature and exceptional color rendering index [1–4]. In addition, the performances of n-UV LEDs strongly depend on the luminescence performance of phosphors used. Therefore, the recent researches have focused on developing suitable tricolor phosphors with high luminescent efficiency and high stability for n-UV LEDs [4–6].

As we all know, Tb³⁺ ion is frequently used as an activator of green-emitting luminescent phosphors because it has a relatively simple structure of energy levels that consist of ⁷F _J , ⁵D₄, and ⁵D₃ states. Usually, the ⁵D₄–⁷F₅ transition peaking at around 541 nm leads to the predominant green emission, which is a magnetic dipole transition with $\Delta J\; = \;1$ [9, 34]. However, the intensities of Tb³⁺ characteristic sharp lines at 488, 541, and 582 nm are very week and the spectral widths are quite narrow due to the strictly forbidden f–f absorption transitions as well. One of the efficient strategies to conquer the difficulty is to introduce Ce³⁺ as a sensitizer to transfer excitation energy to Tb³⁺ efficiently because there is larger spectral overlapping between the emission spectrum of Ce³⁺ ion and absorption spectrum of Tb³⁺ ion. Until now, several Ce³⁺/Tb³⁺ co-doped phosphors have been synthesized and investigated, such as Li₆Lu(BO₃)₃:Ce³⁺, Tb³⁺; Y₄Si₂O₇N₂:Ce³⁺, Tb³⁺; Ca₆Ba(PO₄)₄O:Ce³⁺, Tb³⁺; YBO₃:Ce³⁺, Tb³⁺; KGdF₄:Ce³⁺, Tb³⁺; GdPO₄:Ce³⁺, Tb³⁺, and so on [3–20].

Materials belonging to the large apatite family described by the general formula of M₁₀(TO₄)₆X₂ (M = alkaline; X = monovalent anion, such as OH⁻ Cl⁻, and F⁻; and TO₄ ³⁻ = trivalent anions, such as PO₄ ³⁻, VO₄ ³⁻, or AsO₄ ³⁻) are widely used for n-UV LEDs owing to their many advantages including high chemical and physical stability, low cost, and excellent weather resistance [21–25]. A wide range of cationic and anionic substitutions are possible to create many modified apatite structures with improved luminescent performance [25]. Especially, Ca₁₀(PO₄)₆Cl₂ is a good host lattice for luminescence. Recent investigation was conducted on Ca₁₀(PO₄)₆Cl₂:Eu³⁺; Ca₁₀(PO₄)₆Cl₂:Eu²⁺ [21, 23]. To the best of our knowledge, there is no report on the luminescence of Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors yet. Herein, we prepared Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors via a high-temperature solid-state reaction method. The crystal structure, photoluminescent (PL) properties, color chromaticity, energy transfer, and thermal quenching of the as-prepared samples were investigated. The results show that Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors could be a potential blue-green tunable phosphor for n-UV pumped WLEDs.

Experimental section

Synthesis

Ca₉Sr(PO₄)₆Cl₂:x%Ce³⁺ ($1\; \le \;x\; \le \;7$) and Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, y%Tb³⁺ ($1\; \le \;y\; \le \;7$) phosphors were prepared by a solid-state reaction from a stoichiometric mixture of CaCO₃, SrCl₂, (NH₄)₂HPO₄, CeO₂, and Tb₄O₇. The weighed raw chemicals were ground thoroughly in an agate mortar for 2.5 h, and then the obtained mixture was sintered in alumina crucibles at 1400 °C for 5 h under CO reducing atmosphere to produce the final sample. Finally, the prepared samples were cooled to room temperature and reground for further measurements.

Characterizations

To investigate the crystal phases of the as-prepared powder, X-ray diffractions (XRD) were performed in the range of $10\; \le \;2\theta \; \le \;70^{ \circ }$ with the step of 0.02^o using a Shimadzu-6000 X-ray generator equipped with Cu $K_{\alpha }$ radiation (λ = 0.15406 nm). Diffuse reflectance spectra were measured using a Shimadzu UV-2600 spectrophotometer using the white BaSO₄ powder as a reference standard. Photoluminescence emission (PL) and excitation (PLE) were recorded on an F-280 fluorescence spectrophotometer with a 150 W Xe lamp as its excitation source. A homemade temperature control system was used to measure temperature-dependent emission spectra from 298 to 573 K, in which the measuring and controlling accuracy of the temperature is about ± 0.5 °C.

Results and discussion

Phase and crystal structure

The XRD patterns of Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺; Ca₉Sr(PO₄)₆Cl₂:5%Tb³⁺; and Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, 5%Tb³⁺ samples are presented in Fig. 1. It can be found that all XRD patterns have the similar profiles, which are well consistent with that of the standard hexagonal Ca₁₀(PO₄)₆Cl₂ reported in ICSD card with the number of 24237 (space group P63/m, no.176). No detectable phase from impurity can be observed in all the as-prepared samples even though Sr²⁺, Ce³⁺, and Tb³⁺ ions were introduced into the host lattice. This suggests that all of the samples are of single phase and the Sr²⁺, Ce³⁺, and Tb³⁺ ions have been successfully incorporated in the Ca₁₀(PO₄)₆Cl₂ host lattice without changing the crystal structure. In addition, we can also see that the XRD peaks of the as-prepared samples are intense and sharp, indicating that the products are crystallized well. This is in the favor of highly efficient luminescence of rare earth ions as well.

Fig. 1

Typical XRD patterns of Ca₉Sr(PO₄)₆Cl₂:5mol%Ce³⁺, Ca₉Sr(PO₄)₆Cl₂:5mol%Tb³⁺, and Ca₉Sr(PO₄)₆Cl₂:5mol%Ce³⁺, 5mol%Tb³⁺ phosphors as well as the ICSD card of the standard Ca₁₀(PO₄)₆Cl₂ crystal

Diffuse reflectance spectra

Figure 2 shows the diffuse reflectance spectra of Ca₉Sr(PO₄)₆Cl₂, Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, Ca₉Sr(PO₄)₆Cl₂:5%Tb³⁺, and Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, 5%Tb³⁺ phosphors, which are good evidence of absorption in the n-UV region induced by the activator ions. For Ca₉Sr(PO₄)₆Cl₂ host, it can be found that a very weak absorption band ranges from 230 to 420 nm. When Ce³⁺ ions are introduced into the host, a strong and broad absorption band in the range of 230–500 nm can be observed, which is ascribed to the 4f–5d absorption of Ce³⁺ ion. For Ca₉Sr(PO₄)₆Cl₂:5%Tb³⁺, the spectral profile is similar to that of the undoped host except a strong absorption peak at around 250 nm, which corresponds to the 4f-5d transition of Tb³⁺ ion [12]. However, as Ce³⁺ and Tb³⁺ ions are introduced into the host together, the intensity of absorption band of Ce³⁺ ions are improved obviously, reflecting the occurrence of an energy transfer from Ce³⁺ to Tb³⁺. Therefore, the above results also suggest that Ce³⁺–Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphor can be used as a blue-green tunable phosphor for n-UV LED because of an efficient absorption in the n-UV region.

Fig. 2

The diffuse reflectance spectra of Ca₉Sr(PO₄)₆Cl₂:x mol%Ce³⁺, y mol%Tb³⁺ phosphors

Photoluminescence properties and concentration quenching of Ce³⁺ ions

Figure 3a shows the photoluminescent emission (PL) and excitation (PLE) spectra of the Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺ phosphor. It can be found that the PL spectrum shows a broad nonsymmetric blue emission band extending from 360 to 700 nm with a peak centered at about 431 nm. Since there are two Ca²⁺ sites in the Ca₉Sr(PO₄)₆Cl₂ host and Ce³⁺ usually presents typical double emission peaks in a single definite lattice site from the lowest 5d excited state to the ²F _J (J = 5/2, 7/2) spin–orbit split 4f ground states, the emission spectrum can be decomposed into four Gaussian components with peaks centered at 387, 418, 444, and 487 nm, which are labeled as band A, B, C, and D, respectively [9, 21, 23]. The energy gaps between A and B, and C and D are 1950 and 1989 cm⁻¹, respectively, which is in good agreement with the theoretical value of about 2000 cm⁻¹ [27, 28]. From the emission spectrum, therefore, we concluded that there are two kinds of Ce³⁺ luminescent centers in the Ca₉Sr(PO₄)₆Cl₂ host lattice. Monitoring the emission at 431 nm, the PLE spectrum exhibits a broad band in the range from 220 to 385 nm, which is derived from the 4f–5d transition of Ce³⁺ ion. The strongest absorption band peaked at about 324 nm, which matches well with n-UV LED chips.

Fig. 3

a PL and PLE spectra of Ca₉Sr(PO₄)₆Cl₂:5 mol % Ce³⁺ phosphor and four Gaussian fitting emission spectra. b Normalized PL spectra of Ca₉Sr(PO₄)₆Cl₂:xmol%Ce³⁺ phosphors with different Ce³⁺ doping concentrations. The inset shows the dependence of luminescent intensity on the doping concentration of Ce³⁺ ions

In order to investigate the effect of doping concentration on the luminescent properties, a series of Ca₉Sr(PO₄)₆Cl₂:x%Ce³⁺ ($1 \le \;x\; \le \;7$) phosphors were synthesized. The inset of Fig. 3b shows the dependence of emission intensity on the concentrations of Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂ phosphors. It can be observed that the blue emission of the Ce³⁺ gradually increases with increasing Ce³⁺ concentration, and reaches a maximum value at x = 5. Continuing the doping concentration of Ce³⁺ ions, we can find that emission intensity decreases because of the concentration quenching effect. In the meantime, an obvious redshift of the emission peak wavelength is observed in Fig. 3b as the concentration of Ce³⁺ gradually increases. This phenomenon can be explained as follows. In our case, we propose that there exists an energy transfer between Ce³⁺ ions in different crystallographic sites. Thus, the intensities of four Gaussian components (Ce_A–Ce_D) originated from two kinds of Ce³⁺ increased relatively with increasing Ce³⁺ concentration, and it affected the change of emission shape and the peak position by the substitution of Ca²⁺ ions [6, 11, 28]. Furthermore, the reabsorption of the high-energy part of the emission (resonant with the low-energy part of the excitation spectra) is another possible reason for the redshift of the emission spectrum in Ca₉Sr(PO₄)₆Cl₂:x%Ce³⁺ ($1 \le \;x\; \le \;7$) phosphors [29].

Usually, it is considered that concentration quenching is mainly caused by energy transfer among Ce³⁺ ions. Therefore, to further investigate the energy transfer mechanism of Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂, the critical distance of energy transfer (R _c) should be evaluated first. According to the theory proposed by Blasse, the critical distance, R _c, can be expressed as follows [29, 30]:

$$ R_{\text{c}} \approx \;2\left[ {\frac{3V}{{4\pi x_{\text{c}} N}}} \right]^{1/3}, $$

(1)

where V is the volume of the unit cell, x _C is the critical concentration of Ce³⁺ ions, and N is the number of host cations in the unit cell. As for Ca₉Sr(PO₄)₆Cl₂ host, V = 537.64 Å³, x _C = 0.05, and N = 10. Therefore, the critical transfer distance R _c is calculated to be 12.71 Å. Non-radiative energy transfer may occur by exchange interaction or multi-polar interaction. The former one is dominant when the critical distance is <4 Å [30, 31]. So it can be excluded that the exchange interaction is responsible for the energy transfer between Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂ host. Therefore, electric multipole should be the dominant mechanism for the energy transfer between Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂:Ce³⁺ phosphors. Van Uitert has developed a phenomenological model to explain the relationship between the luminescent intensity and the concentration of luminescent center, which can be written as [32, 33]

$$ \frac{I}{x} = k[1 + \beta (x)^{\theta /3} ]^{ - 1}, $$

(2)

where x is the doping concentration of Ce³⁺ ions in the present case; K and β are constants for a certain system; θ represents the interaction type between luminescence center and quenching center, here θ = 6, 8, or 10, indicating the exchange interaction, electric dipole–dipole (D–D), electric dipole–quadrupole (D–Q), and electric quadrupole–quadrupole (Q–Q) interactions, respectively. In order to understand the energy transfer mechanism between Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂ phosphors, Eq. (2) was used to fit the experimental data in Fig. 4. It can be found that Eq. (2) fit well with the experimental data, and the θ value was deduced from the fitting process to be 6.21, which suggests that the electric dipole–dipole interaction should be responsible for the energy transfer of Ce³⁺ in Ca₉Sr(PO₄)₆Cl₂ phosphors.

Fig. 4

Dependence of integrated emission intensity of Ce³⁺ ions on doping concentration in the Ca₉Sr(PO₄)₆Cl₂:Ce³⁺ phosphors. Solid curve is the fitting curve derived based on Van Uitert’s model

Energy transfer from Ce³⁺ to Tb³⁺

As for the Tb³⁺-doped Ca₉Sr(PO₄)₆Cl₂ sample, its PLE and PL spectra are presented in Fig. 5b. The PL spectrum under the excitation of 379 nm displays a series of sharp-line emissions. The emission peaks at 486, 541, 590, and 618 nm are assigned to the ⁵D₄–⁷F _J (J = 6, 5, 4, 3) characteristic transitions of Tb³⁺ ion. The green emission line at 541 nm from the ⁵D₄–⁷F₅ transitions dominates the whole emission spectrum, which is a magnetic dipole transition with $\Delta J\; = \;1$ [9, 34]. Therefore, the Tb³⁺-doped Ca₉Sr(PO₄)₆Cl₂ sample exhibits green emission under the excitation of 379 nm. Monitoring the emission at 541 nm, the PLE spectrum contains a broad band and several lines. The broad band ranging from 200 to 270 nm centered at 256 nm is assigned to the 4f–5d transition of Tb³⁺ ions, while the excitation lines are attributed to the intra-4f ⁸ transitions of Tb³⁺ ions [2]. Comparing the PL spectra of Ce³⁺ (Fig. 5a) with the PLE spectra of Tb³⁺ (Fig. 5b), it is clearly observed that there is a spectral overlap, indicating that an effective energy transfer from the sensitizer, Ce³⁺, to the activator, Tb³⁺, can be expected in the Ca₉Sr(PO₄)₆Cl₂ host. Meanwhile, as shown in Fig. 5c, it can be found that the profile of PLE spectrum of Ce³⁺–Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphor monitored the emission of Tb³⁺ ions at 541 nm is similar to that monitored the emission of Ce³⁺ ions at 431 nm. Moreover, when the sample was excited by 324 nm UV light, the PL spectrum shows the weaker blue emission of Ce³⁺ and the stronger green emission of Tb³⁺ comparing with that in Ca₉Sr(PO₄)₆Cl₂:Ce³⁺ and Ca₉Sr(PO₄)₆Cl₂:Tb³⁺, which is an obvious evidence for the energy transfer from Ce³⁺ to Tb³⁺ in Ca₉Sr(PO₄)₆Cl₂ host [35]. The PLE spectra consist of a broad band extending from 250 to 360 nm which means that Ca₉Sr(PO ⁾_{4 6}Cl₂:Ce³⁺,Tb³⁺ phosphor is a potential green-emitting phosphor for UV pumped WLEDs as well.

Fig. 5

PLE and PL spectra of Ca₉Sr(PO₄)₆Cl₂:Ce³⁺, Tb³⁺ phosphors. a 5 mol% Ce³⁺, b 5 mol%Tb³⁺, and c 5 mol% Ce³⁺, 5 mol% Tb³⁺

Figure 6 shows the PL spectra of the as-prepared Ca₉Sr(PO₄)₆Cl₂:5%molCe³⁺, y%molTb³⁺ (0 ≤ y ≤ 7) samples under the excitation of 324 nm. It can be seen that all the emission spectra mainly consist of a broad emission band of several line peaks. This form corresponds to the 4f–5d transitions of Ce³⁺ ion and the later originates from the ⁵D₄–⁷F _J (J = 6, 5, 4, 3) characteristic transitions of Tb³⁺ ion. Moreover, the emission intensity of the Ce³⁺ decreases monotonously with increasing the Tb³⁺ concentration when the Ce³⁺ doping concentration is fixed, which indicates an efficient energy transfer from Ce³⁺ to Tb³⁺. From the luminescence of Tb³⁺ ions, it can be found that the emission intensity of Tb³⁺ increases gradually and reaches a maximum value when y = 5. As a result of concentration quenching, the emissions of Tb³⁺ decreases when the Tb³⁺ ion concentration exceeds over 5 %.

Fig. 6

PL spectra of Ca₉Sr(PO₄)₆Cl₂:5mol% Ce³⁺, ymol%Tb³⁺ (0 ≤ y ≤ 7) phosphors under the excitation of 324 nm

In general, the energy transfer efficiency from a sensitizer to activator, η _ET, can be expressed as the following equation [36]:

$$ \eta_{\text{ET}} = 1 - \frac{{I_{\text{s}} }}{{I_{{{\text{s}}0}} }}, $$

(3)

where I _s is the luminescent intensity of Ce³⁺ in the presence of Tb³⁺ and I _s0 is the luminescent intensity of Ce³⁺ in the absence of Tb³⁺. In the Ca₉Sr(PO₄)₆Cl₂:Ce³⁺, Tb³⁺ system, Ce³⁺ is the sensitizer and Tb³⁺ is the activator. Figure 7 shows the result of energy transfer efficiency from Ce³⁺ to Tb³⁺ calculated using Eq. (3). As shown in Fig. 7, the energy transfer efficiency increases with the increase of Tb³⁺ concentration. However, the increscent rate of the emission intensity gradually decreases with increasing Tb³⁺ concentration. The maximal energy transfer efficiency can reach 76 % when 7 %mol Tb³⁺ ions were doped. The above results indicate that the energy transfer from Ce³⁺ to Tb³⁺ is very efficient.

Fig. 7

Dependence of the energy transfer efficiency ($\eta_{\text{T}}$) from Ce³⁺ to Tb³⁺ on the Tb³⁺ doping concentration in the as-prepared Ca₉Sr(PO₄)₆Cl₂ phosphors

On the basis of Dexter’s energy transfer formula of multi-polar interaction and Reisfeld’s approximation, the following relation can be obtained [34–37]:

$$ \frac{{\eta_{0} }}{\eta } \propto C^{n/3}, $$

(4)

where $\eta_{0}$ and $\eta$ are the luminescent quantum efficiency of the sensitizer (Ce³⁺) in the absence and presence of an activator (Tb³⁺), respectively and C is the concentration of Tb³⁺ ions. The value of $\eta_{0} /\eta$ was estimated by the ratio of luminescence intensities as [36–38]

$$ \frac{{I_{0} }}{{I_{\text{s}} }}\,{ \propto }\,C^{n/3}, $$

(5)

where I ₀ and I _S are the intrinsic luminescence intensity of a sensitizer (Ce³⁺) in the absence and presence of an activator (Tb³⁺), respectively; n = 6, 8, and 10 corresponding to dipole–dipole (d–d), dipole–quadrupole (d–q), and quadrupole–quadrupole (q–q) interactions, respectively. The I _S0 /I _S ~ C ^n/3 plots are illustrated in Fig. 8. Best linear behavior can be observed only when n = 8, indicating that energy transfer from Ce³⁺ to Tb³⁺ occurred via the electric dipole–quadrupole interaction (d–q) mechanism in the Ca₉Sr(PO₄)₆Cl₂ host.

Fig. 8

Dependence of I/I₀ of Ce³⁺ ions on a C^6/3, b C^8/3, and c C^10/3

Color chromaticity

Because the spectral components can be greatly changed with the increase of Tb³⁺ concentration, as a result, the PL emitting colors can be tuned. In order to more intuitionistically observe the effect of Ce³⁺ doping concentration on PL emission colors, the Commission International del’Eclairage (CIE) chromaticity coordinates of Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, y%Tb³⁺ (0 ≤ y ≤ 7) phosphors were calculated and the typical results are shown in Table 1 and Fig. 9. It can be seen that the chromaticity coordinates tune from (0.169, 0.151) to (0.258, 0.460) with increasing the concentration of Tb³⁺ from 0 to 7 %. Correspondingly, the color tone of the phosphors can be adjusted from indigo to green under the excitation of UV light, which suggests that the phosphor may have potential application for n-UV pumped WLEDs.

Table 1 CIE chromaticity coordinates of Ca₉Sr(PO₄)₆Cl₂:5mol%Ce³⁺, ymol%Tb³⁺ (0 ≤ y ≤ 7) phosphors under the excitation of 324 nm

Fig. 9

CIE chromaticity diagram of Ca₉Sr(PO₄)₆Cl₂:5mol%Ce³⁺, ymol%Tb³⁺ (0 ≤ y ≤ 7) phosphors under the excitation of 324 nm

Thermal quenching

For the application of high-power WLEDs, thermal stability is one of the most important parameters for phosphors, because it can considerably influence the light output, color rendering index, and stability of devices. Herein, temperature-dependent emission spectra of the selected Ca₉Sr(PO₄)₆Cl₂:5%Ce³⁺, 5%Tb³⁺ phosphor at different temperatures from 298 to 573 K under the excitation of 324 nm are shown in Fig. 10a. It is obvious that the emission intensities of the sample decrease with increasing temperature from 298 to 573 K. Especially, the PL intensity at 150 °C drops to be around 57.6 % of the initial value at room temperature, which suggests that the as-prepared phosphors possesses good thermal stability properties (as shown in the inset of Fig. 10a). Generally, the decrease of emission intensity is ascribed to the thermal quenching of emission intensity via phonon interaction, in which the excited luminescent center is thermally activated through the crossing point between the ground and the excited states [39, 40]. In order to calculate the activation energy (ΔE) for thermal quenching and to better understand the thermal quenching process, the temperature-dependent emission intensity is described by a modified Arrhenius equation [6, 41, 42]:

Fig. 10

aThe PL spectra ($\lambda_{\text{ex}} = 324{\text{ nm}}$) of Ca₉Sr(PO₄)₆Cl₂:5mol%Ce³⁺, 5mol%Tb³⁺ phosphor at different temperatures from 298 K to 573 K. The inset shows the relative intensity as a function of temperature of the selected phosphor. b A fitting line of ln[(I ₀ /I)−1] vs.1/kT activation energy graph for thermal quenching of the corresponding sample

$$ I_{T} = \frac{{I_{0} }}{{1 + c\exp \left( {\frac{ - \Delta E}{kT}} \right)}}. $$

(6)

Herein, I(T) and I ₀ are the intensities of the initial and different temperatures, respectively; c is a constant; k is the Boltzmann’s constant (8.617 × 10⁻⁵ eV); and ΔE is the activation energy for thermal quenching. So Eq. (6) can be rewritten as

$$ \ln \left( {\frac{{I_{0} }}{I} - 1} \right) = \ln c - \frac{\Delta E}{kT}. $$

(7)

According to Eq. (7), the activation energy ΔE can be calculated from a plot of ln[(I ₀ /I)−1] against 1/kT. As shown in Fig. 10b, a linear fitting can be obtained using Eq. (7). The slope of the linear line equals −ΔE. Thus, the ΔE was deduced to be 0.176 eV.

Conclusions

To sum up, a series of Ce³⁺-doped and Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphors have been prepared via solid-state reaction successfully. The optimal doping concentration of Ce³⁺ ions in Ca₉Sr(PO₄)₆Cl₂ phosphor was confirmed to be 5 mol%. Moreover, an obvious redshift of emission band was observed with the increase of Ce³⁺. For the Ce³⁺/Tb³⁺ co-doped Ca₉Sr(PO₄)₆Cl₂ phosphor, the emission intensity of Ce³⁺ ions dramatically decreases with the increase of Tb³⁺ ion concentration because of an energy transfer from Ce³⁺ to Tb³⁺. Moreover, the energy transfer efficiency increases greatly with the increase of Tb³⁺ ion concentration. The mechanism of energy transfer between Ce³⁺ and Tb³⁺ was deduced to be electric dipole–dipole–quadrupole interaction. The results of temperature-dependent emission spectra indicate that the Ca₉Sr(PO₄)₆Cl₂: Ce³⁺,Tb³⁺ phosphor has excellent thermal stability of luminescence. The results above suggest that the as-prepared phosphors can be used as blue and green components for n-UV pumped WLEDs.