1. Introduction

Long persistent phosphor (LPP) is a kind of materials which can store the absorbed light energy and then release it gradually for an appreciable time in the form of luminescence after the removal of the excitation light source [1–3]. The phenomenon of the seeming “self-sustained” luminescence is generally called long persistent luminescence (LPL). Due to the environmentally friendly and energy-saving properties, LPPs have attracted great attentions. They can be applied in various important fields extending from night displays, security signs, optical storage media, fiber-optic thermometer, medical diagnostics to in vivo bio-imaging [4–7].

Although a large amount of efforts have spent on the mechanistic studies, the mechanism of LPL has not yet been reached a consensus. Some minor but critical details such as the formation, distribution and electronic structure of traps in the host are still remains an open question. In addition, there is few solid experimental evidence to support those existed assumptions on the LPL mechanism [8, 9]. At present, the mechanism of LPL generation is considered as the recombination of the electrons released from traps with the emission centers. Naturally, the suitable trap depth and high density play decisive roles in determining the LPL duration and brightness. To design a LPP, the foremost thing is to select a proper host material and emission center. Second, the defects working as suitable trap centers are needed. Usually, the intrinsic traps such as oxygen vacancies and cation vacancies can be created in the host lattice automatically under higher temperature sintering environment and reducing atmosphere [10–12]. Besides, the extrinsic traps can be introduced artificially by doping or codoping with aliovalent ions [13–16] though the accurate trap depth is hard to be controlled.

Ce³⁺ is widely used as an activator or sensitizer in inorganic materials in view of their potential applications as scintillators and phosphors. Ce³⁺-doped phosphors yield broad band emission with different colors due to parity- and spin-allowed 5d-4f transitions, which strongly relies on the strength of the crystal field of the host [17, 18]. It may be a potential candidate of activator for the synthesis of LPPs. Moreover, Ce³⁺ is the simplest ion among trivalent rare earth ions since that it just has one electron of ground state (4f¹) and excited state (5d¹), respectively, which is helpful to explore and study the LPL mechanism. However, among the developed LPPs, the study of Ce³⁺-doped LPP is comparatively rare in spite of the early reported Ce³⁺-doped LPPs including Ca₃Al₂O₆:Ce³⁺ [19], Ca(Ba)Al₂O₄:Ce³⁺ [20, 21], Ca₂Al₂SiO₇:Ce³⁺ and CaYAl₃O₇:Ce³⁺ [22], and some newly reported Ce³⁺-doped LPPs based on garnet-type hosts [23–26]. Eu²⁺-doped LPPs based on silicates have been reported [27].Therefore, it is of great interest to synthesize and investigate the novel Ce³⁺-doped LPPs based on silicates.

In this paper, a novel blue LPP of Ce³⁺- doped Na₂Ca₃Si₂O₈ was prepared successfully via the traditional high temperature solid-state reaction method. The phase structure and photoluminescence (PL) properties for Na₂Ca₃Si₂O₈: Ce³⁺ were investigated. As previously expected, the blue LPL was observed. The LPL decay curves and thermoluminescence (TL) results were discussed in detail. The trap types and dynamics were illustrated. Based on the results, a possible mechanism of LPL generated in for Na₂Ca₃Si₂O₈: Ce³⁺ was also described.

2. Experimental

2.1 Synthesis

A series of crystalline powder samples of Ce³⁺ solely-doped Na₂Ca₃Si₂O₈ were synthesized successfully by a conventional high temperature solid-state reaction. CaCO₃ (99.0%), SiO₂ (99.0%), Na₂CO₃ (99.8%) and CeO₂ (99.99%) were used as raw materials. The starting materials were weighed according to the nominal compositions of Na₂Ca_3-xSi₂O₈: xCe³⁺ (x = 0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06 and 0.07). Then, the mixtures were wet ground thoroughly in an agate mortar with alcohol as the wetting medium. After that, the mixtures were calcined at 1300 °C for 5 h in a tube furnace in reducing atmosphere (90% N₂ + 10% H₂). Finally, the samples were furnace-cooled to room temperature and ground again into fine powders for the following measurements.

2.1 Measurement

The phase structure of all as-prepared powder samples were characterized by a XD-2 powder diffractometer (Beijing PGENERAL) using Cu Kα irradiation (λ = 1.5406 Å) at 36 kV tube voltage and 20 mA tube current with a continuous scanning step of 0.02° in the 2θ range from 10° to 70°. Diffuse reflectance spectrum was measured by a UV-2450 UV-Visible spectrophotometer (Japan) using BaSO₄ as a reference. Room temperature photoluminescence excitation (PLE) and PL spectra were measured by a Hitachi F-7000 fluorescence spectrophotometer with a photomultiplier tube operating at 400 V, and a Xe lamp (150 W) used as the excitation source. Suitable optical filters were used to eliminate the interference of diffraction peaks. The LPL spectrum and decay curves were also recorded on the same fluorescence spectrophotometer. Before the measurement of the LPL spectrum and decay curves, the samples were first irradiated for 1 min with a 254 nm UV light source. TL glow curves were measured by using a FJ427A1 thermoluminescent dosimeter (CNNC Beijing Nuclear Instrument Factory) form room temperature to 360 °C with a linear heating rate of 1°C /s.

3. Results and discussion

3.1 Phase structure

Figure 1 displays the XRD patterns of a representative sample Na₂Ca_2.95Si₂O₈: 0.05Ce³⁺, thereinto, the diffraction peaks are well indexed by the standard data of Na₂Ca₃Si₂O₈ (JCPDF standard card No.23-0668) which belongs to triclinic structure. The XRD patterns show no impurity peaks, which infers that a single-phased structure is formed. On the basis of the closer ionic radius and valence of the Ce³⁺ ion (101 pm) and Ca²⁺ ion (99 pm), it is expected that Ce³⁺ ions probably substitute for the Ca sites.

 

Fig. 1 XRD patterns of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ and that of standard data of Na₂Ca₃Si₂O₈ (JCPDS 23-0668) given for comparison.

Download Full Size | PDF

3.2 UV-Vis Di use reflectance spectra

Figure 2(a) presents the diffuse reflectance spectra of samples Na₂Ca₃Si₂O₈. The reflectance spectrum of the sample shows platform of high reflection in the wavelength range of 350–800 nm. The reflectance spectra of Na₂Ca₃Si₂O₈ host shows an absorption band before 350 nm, which is attributed to the host absorption. According to Tauc [28, 29], The band gap is estimated to be about 4.30 eV as shown in Fig. 2(b).

 

Fig. 2 (a)Diffuse reflectance spectrum of sample Na₂Ca₃Si₂O₈ ; (b) The function of [F(R)hλ]² versus photon energy hλ according to Tauc.

Download Full Size | PDF

3.3 Photoluminescence and persistent luminescence properties

Figure 3(a) shows the PLE and PL spectra of the typical sample Na₂Ca_2.97Si₂O₈: 0.03Ce³⁺. As shown in Fig. 3(a), four broad PLE bands are obviously observed at ~336nm, 288nm, 250nm and 222nm by monitoring the emission at 401 nm. The PLE bands at 336nm, 288nm, 250nm and 222nm can be assigned to the transitions from the ground energy level to excited levels of Ce³⁺ 5d₁, 5d_2, 5d₃ and 5d₄, which is similar to that of Ce³⁺ in Li₂CaGeO₄ [30] and ABaPO₄ (A = Li, Na, K) [31]. The excitation spectrum of Ce³⁺ directly shows the splitting of 5d orbital into four levels mentioned above in the crystal field. Under excitation of 336nm, Na₂Ca_2.97Si₂O₈: 0.03Ce³⁺ shows blue emission with a broad unsymmetrical doublet bands peaking at 401nm extending from 340 to 530nm. In general, the doublet emission character of Ce³⁺ is caused by transitions from the lowest 5d excited states to the spin–orbit splitting of ground states (²F_5/2 and ²F_7/2). As shown in Fig. 3(b), the asymmetric emission band can be well fitted to two Gaussian profiles centered at 24160 cm⁻¹ (414 nm) and 25904 cm⁻¹ (386 nm). Considering the strong coupling with host lattice, the energy difference between two emission bands is about 1744 cm⁻¹ which is comparable to the theoretical value (ca. 2000 cm⁻¹) for ²F_5/2 and ²F_7/2 ground states of Ce³⁺. Herein, the Stokes shift of Na₂Ca₃Si₂O₈:Ce³⁺ was estimated approximately to be 5272 cm⁻¹. Figure 3(c) gives the emission intensity of Ce³⁺ as a function of doping concentration. It can be noticed that the emission intensity of Ce³⁺ increases with the rising of its concentration, reaching the maximum at x = 0.03, and then decreases with further increasing doping concentration. It is caused by the concentration quenching effect and the optimum doping concentration of Ce³⁺ in Na₂Ca₃Si₂O₈ host is experimentally determined to be 0.03. The concentration quenching effect is mainly caused by energy transfer among Ce³⁺ ions. So, the close distance between Ce³⁺ facilitated by increasing doing of the Ce³⁺ ions plays an important role.

 

Fig. 3 (a) The PLE and PL spectra of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺; (b) The fitted result of the unsymmetrical doublet band for the typical sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺; (c) The relative emission intensity of Ce³⁺ versus doping concentrations.

Download Full Size | PDF

The LPL can be observed after the irradiation by UV light (200-340 nm). Figure 4 shows the LPL spectrum of sample Na₂Ca_2.98Si₂O₈:0.02Ce³⁺ after the 254 nm excitation source was switched off. It reveals one broad emission band which is almost the same as the PL spectrum in terms of emission band shape and position, indicating that the LPL originates from the same emission center of Ce³⁺. Thus, the blue LPL phenomenon can be observed as exhibited in the inset of Fig. 4.

 

Fig. 4 LPL spectrum of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ after the removal of UV light source. Inset: the digital photo of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ after stopping UV irradiation.

Download Full Size | PDF

3.4 Persistent luminescence decay curves

Figure 5 shows the LPL decay curves of the phosphors Na₂Ca_3-xSi₂O₈:xCe³⁺ (x = 0.01-0.07) at room temperature in the time range of 0-1000s. Notably, the LPL decreases rapidly at first and then decays very slowly. Owning to the slow decay process, the LPL emission can be observed. The doping concentration of Ce³⁺ has great effect on the LPL decay behaviors. The slow decay component can last the longest time when x = 0.03. Lower and higher doping concentration of Ce³⁺ markedly decreases the LPL duration. It indicates the optimal concentration of Ce³⁺ for LPL is about 0.03.

 

Fig. 5 LPL decay curves of samples Na₂Ca_3-xSi₂O₈:xCe³⁺ (x = 0.01-0.07).

Download Full Size | PDF

3.5 Thermoluminescence properties

It is widely recognized that the traps in the host lattice play a crucial role in LPPs. The energy stored in too shallow traps will be released quickly and in deep traps may not be released at room temperature. Both cases are not beneficial for the good LPL properties. It is commonly assumed that the ideal trap is situated slightly above room temperature 50-120 °C [2, 10, 32, 33]. In order to investigate the generation of LPL, the knowledge of the traps state, the motion of charge carriers in the host lattice are must be studied. Currently, it is generally accepted that the most efficient tool applied to study the LPL properties is thermoluminescece technique. In order to get further information about the trapping centers and further investigate the significant effect of different traps on the LPL properties, the measurement of TL glow curves of Ce³⁺-doped Na₂Ca₃Si₂O₈ samples were performed.

Figure 6(a) presents the results of TL measurements for non-doped and Ce³⁺-doped Na₂Ca₃Si₂O₈, hence, changing the concentration of Ce³⁺. It can be seen that the intensity of TL curve is largely relied on the doping concentration of Ce³⁺. TL curves of all obtained samples appear as a broad band in the temperature range of 35-350 °C. Apparently, each TL glow curve is not a single symmetry band and may be composed of several overlapping peaks. Therefore, Fig. 6(b) and 6(c) show the TL curves of samples Na₂Ca_3-xSi₂O₈:xCe³⁺ (x = 0 and 0.03) separately and the deconvoluted results by curve fitting technique based on the Gaussian equation. As shown in Fig. 6(b), the TL curve of non-doped samples can be deconvoluted into three components: band a (147 °C), band b (200 °C) and band c (300 °C) which are probably caused by intrinsic defects such as oxygen vacancies. Similarly, Fig. 6(c) shows the fitting results of three Gaussian bands for Ce³⁺-doped samples: band 1 (114 °C), band 2 (193 °C) and band c (304 °C). There are two main significant differences in TL curves between non-doped and Ce³⁺-doped Na₂Ca_3-xSi₂O₈ samples. Firstly, though the positions of bands 2 and 3 are close to that of bands a and b, which can be attributed to one kind of traps, respectively, the large deviation of band 1 and band a may infers that the new traps are created by the incorporation of Ce³⁺ into Na₂Ca₃Si₂O₈ host lattice and the intrinsic traps at band a are buried in new traps. Secondly, with the doing of Ce³⁺ ions, the intensity of TL band becomes hundreds of times stronger than that of non-doped one. It infers that the incorporation of Ce³⁺ also enhances the generation of intrinsic traps to a large extent.

 

Fig. 6 (a) TL glow curves of samples Na₂Ca_3-xSi₂O₈:xCe³⁺ (x = 0-0.07); The fitting results of samples Na₂Ca₃Si₂O₈ (b) and Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ (c) based on the Gaussian function.

Download Full Size | PDF

In order to obtain some indications on the number TL bands of the Ce³⁺-doped TL curves and their individual temperatures, a procedure of the peak resolution suggested by McKeever [34] was carried out. Figure 7(a) presents the TL curves of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ after a short preheating of the specimen to increasing different temperatures (T_Stop). Based on it, the T_M-T_Stop (T_M is the temperature at TL band maximum) dependence is potted in Fig. 7(b). It is noted that a “staircase” shaped T_M-T_Stop curve, in the temperature range of 70-130 °C, where a flat area can be distinguished. According to McKeever, the plateau corresponding to trap 1 (Fig. 6(c)) is an individual peak. But the traps at higher temperature are more complex. Here, we just investigate the trap 1 quantitatively because that higher temperature traps are too deep to make contribution to the generation of LPL.

 

Fig. 7 (a) TL glow curves of sample Na₂Ca_2.97Si₂O₈:0.03Ce³⁺ collected after preheated at different temperatures; (b) T_M-T_Stop dependence plotted on the basis of the results in (a).

Download Full Size | PDF

In order to estimate the trap depths (E) and trap densities (n₀), Chen’s method was used to analyze the deconvoluted band 1 by the Eqs. (1) and (2):

The estimated E and n₀ corresponding to band 1 are listed in Table 1. The trap depth is estimated to be about 0.45 eV, which indicates one kind of traps is introduced. With the doping of Ce³⁺, the trap density at TL band 1 increases at first and then drops down sharply with a maximum at x = 0.03. When Ce³⁺ is introduced into Na₂Ca₃Si₂O₈ host lattice and substitutes for Ca²⁺ sites, the possible approach to achieve the charge compensation is that two Ce³⁺ ions substituting for three Ca²⁺ ions which result in the formation of two positive charge defects ${\text{Ce}}_{{}^{{\text{Ca}}^{2+}}}^{3+}{}^{•}$ and one negative defect $V_{{}^{{\text{Ca}}^{2+}}}{}^{\prime \prime }$. The defects formation process is expressed as follows: $2{\text{Ce}}^{3+}+3{\text{Ca}}^{2+}\to 2{\text{Ce}}_{{}^{{\text{Ca}}^{2+}}}^{3+}{}^{•}+V_{{\text{Ca}}^{2+}}{}^{\prime \prime }$.

Table 1. Estimated trap depths (E) and density (n₀) of samples Na₂Ca_3-xSi₂O₈: xCe³⁺ (x = 0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06 and 0.07).

View Table

Thus, the TL band 1 can be attributed to the generation of positive charge defects ${\text{Ce}}_{{}^{{\text{Ca}}^{2+}}}^{3+}{}^{•}$ which can serve as electron traps. According to Qu et al. [8], the existence of positive charge defects ${\text{Ce}}_{{}^{{\text{Ca}}^{2+}}}^{3+}{}^{•}$ should assist the formation of $V_{O^{2-}}{{}^{{}^{•}}}^{{}^{•}}$, which confirms assumption that the incorporation of Ce³⁺ ions into host enriches the intrinsic traps $V_{O^{2-}}{{}^{{}^{•}}}^{{}^{•}}$.

3.6 Possible Mechanism of the Na₂Ca₃Si₂O₈:Ce³⁺ phosphors

Based on the results mentioned above, a tentative mechanism by using the energy level scheme was proposed to illustrate the generation of blue LPL in Ce³⁺-doped Na₂Ca₃Si₂O₈ LPPs. Under UV excitation, ground state electrons of Ce³⁺ ions are promoted to the 5d excited states. Some excited electrons will relax to the lowest 5d state and jump to ground levels leading to the 386 and 414 broad band emissions. Besides, when the electrons are pumped to higher 5d states (5d₂ or above), some electrons can be captured by electrons traps with different depths through conduction band (CB) (process ①). When the electrons are pumped to 5d states (5d₁), some electrons can be easily promoted to the CB due to the lower threshold for a thermally assisted photoionization process (process ②). Then, the electrons captured by suitable traps can be moderately released by a thermally stimulated process at room temperature (process ④) and return to 5d state of Ce³⁺ with a series of non-radiation transitions via a motion in CB. The subsequent emission of Ce³⁺ is consistent with that of photoluminescence. Therefore, the blue LPL can be observed in the Na₂Ca₃Si₂O₈:Ce³⁺ LPPs. Those electrons stored in deep traps cannot escape easily at room temperature (process ③), which results in no contribution to LPL (Fig. 8).

 

Fig. 8 Schematic illustration of the persistent luminescence mechanism in Na₂Ca₃Si₂O₈:Ce³⁺.

Download Full Size | PDF

4. Conclusions

In summary, a novel blue-emitting long persistent phosphor Na₂Ca₃Si₂O₈:Ce³⁺ was synthesized successfully by solid-state reaction. The photoluminescence and persistent luminescence properties were characterized. The optimum concentration of Ce³⁺ for photoluminescence and persistent luminescence was experimentally to be about 0.03. Thermoluminescence indicates that there are several stable traps with different depths in this system. The intrinsic and extrinsic traps caused in host lattice were clarified. On the basis of the obtained results, a possible LPL generation mechanism in Na₂Ca₃Si₂O₈:Ce³⁺ is proposed and described.