1. Introduction

Long Persistence Phosphors (LPPs), also known as Long Lasting Phosphors (LLPs), can emit photons for several minutes or hundreds of hours [1] after excitation is stopped. The duration of this persistent luminescence phenomenon can be orders of magnitudes larger than the fluorescence lifetime which is in order of nanosecond or microsecond [2]. For its passive emission features, the LPPs can be used for display devices, emergency beacons, optical storage, radiation detection and so on [3–10]. According to the primary color theory, by adding red, green and blue LPPs together, one can form a color space to reproduce a broad array of colors (a triangle in the CIE 1931 color space). For example, the red-emitting Y₂O₂S:Eu,Ti,Mg [11] and the green-emitting SrAl₂O₄:Eu,Dy [12], together with purplish-blue-emitting CaAl₂O₄:Eu,Dy [13], which have relative high afterglow performance in each color range, can be used for this purpose despite several practical issues such as their different afterglow decay behavior and reabsorption problem.

Physically, persistent luminescence is a thermoluminescence (TL) phenomenon restricted at a specific temperature, the room temperature in convention. This makes the TL technique a frequently-used method in the LPPs study. To analyze TL curves, different methods such as initial rise method and glow peak shape method were proposed. Theories of these two methods are both established on single activation energy physical models [14–19]. According to retrapping rates they treated, these theories are classified as the first, second and general order kinetics [14, 20, 21].

However, when continuous distribution of trap depths is considered, the single activation energy physical models would not be appropriate theoretical bases to construct TL curves analyzing methods [22]. Hornyak et al. built theories for this condition but were specialized to symmetric distributions such as Gaussian distribution or a unit distribution [23].

In practice, though the first, second and general order kinetics can give theoretical curves from asymmetrical to symmetrical, it’s still difficult to decompose a single broad TL peak because shapes of theoretical curves are dependent on their kinetic orders and orders of experimental curves’ components that remain unknown [24–26]. In other words, practical analyzing methods suiting this continuous distribution condition are needed.

As a part of this objective, finding a way to describe the concept of trap depths not confused with those single activation energy models is needed for the study of trap depths continuous distributed LPPs. For this purpose, we applied a technique known as the decay-time method [27] on a bluish-white long persistence phosphor SrSiAl₂N₂O₃:0.001Eu²⁺ to get a series of fading thermoluminescence curves (FTLCs). As revealed by the following experiments, the remarkable trap depths’ continuous distribution property makes this phosphor an excellent probe in the study of the mechanism of LPPs. The crystallography origin of this property is an interesting subject also, but it’s not the topic here. When we subtracted these FTLCs from the least faded one separately, we found almost all the resulting curves had similar valleys at the same temperature, although Urbach's peak position method [15, 22, 28] and the initial rise method showed that each FTLCs expressed different trap depths. This suggests that the temperature obtained by this FTLCs subtracting method does not shift with different fading times and could be regarded as a new characteristic temperature to convey the concept of trap depth in the condition of trap centers continuously distributed. And that this can be checked by just using various FTLCs tests is also an advantage. Fading time is defined as the duration between the end of the light illumination and the beginning of the Thermoluminescence measure in the Thermoluminescence experiment process.

2. Experimental

The sample used in this work was SrSiAl₂N₂O₃:Eu²⁺ which is known as a promising white light-emitting diode phosphor having relatively high thermal stability in a photoluminescence test [29–32]. Structurally, there is only one kind of Sr site in its crystal structure which generally would simplify the discussing process [22, 31]. The persistent luminescence phenomenon of this phosphor has not been reported yet. The sample was prepared by a conventional high-temperature solid state method from raw materials SrCO₃(A.R.), AlN(A.R.), Si₃N₄(A.R.) and Eu₂O₃(99.999%), and 3wt% AlF₃(A.R.) was added as flux. The reaction was performed under an atmosphere of 3NH₃: 10N₂ and a temperature at 1400°C for 5h in a tube furnace.

The phase identification of this sample was tested by powder X-ray diffraction (XRD) analysis using a D2 PHASER X-ray Diffractometer (Bruker Co.) equipped with a graphite monochromator. Its working parameters were λ = 1.54056Å (Cu Kα radiation), voltage = 30kV and current = 15mA. Afterglow (AG) decay curves were recorded using a PR-305 Phosphorphotometer (Hangzhou Zhejiang University Sensing Instruments Co., Ltd) after the samples were irradiated by an artificial sunlight light source with illuminance of 1100lx for 15min at room temperature. Thermoluminescence (TL) curves were measured by a FJ427A Thermoluminescent Dosimeter (CNNC Beijing Nuclear Instrument Factory) with a heating rate of 1K/s and a sample weight of 1.7mg. An 8W, 254nm ultraviolet lamp and an 8W, 365nm ultraviolet lamp were used as illumination source at the same time before the TL measure. Illumination time was 5min.

3. Results and discussion

Figure 1 is the XRD pattern of SrSiAl₂N₂O₃:0.001Eu²⁺, in which no impurity is observed. The relatively higher intensity at 29.0029° and 30.2696° can be ascribed to the rod shaped morphology of crystalline grains [32]. The measured lattice parameters of the sample are a = 4.9502Å, b = 7.9712Å and c = 11.3440Å.

 

Fig. 1 The XRD pattern of SrSiAl₂N₂O₃:0.001Eu²⁺ and calculated x-ray diffraction curve of The Inorganic Crystal Structure Database ICSD#408170.

Download Full Size | PDF

Shown in Fig. 2 is the decay curve of the persistent luminescence of the SrSiAl₂N₂O₃:0.001Eu²⁺ phosphor plotted in a time versus $I^{-1}$ manner. The afterglow persistent time defined by the luminance threshold value 0.32mcd/m² was 2490s for this phosphor. An approximate relationship in the form of $I\propto t^{-1}$ also can be seen in Fig. 2. However, researchers found the $t^{-1}$ decaying behavior of afterglow cannot be explained by the single activation energy physical models based theories [22, 23].

 

Fig. 2 The decay curve of the persistent luminescence of the SrSiAl₂N₂O₃:0.001Eu²⁺ phosphor. Intensity was converted into its reciprocal in order to reveal the relationship between time and afterglow. The sample was illuminated by an artificial sunlight light source with illuminance of 1100lx for 15min at room temperature.

Download Full Size | PDF

By using the Urbach's peak position method, which assigns an energy value (eV) to each TL peak positions (K) via the relation $E_T=T_m/500$ [15, 28], we found trap depths of the FTLGs in Fig. 3 are 0.656eV, 0.658eV, 0.674eV, 0.676eV, 0.692eV, 0.696eV, 0.700eV, 0.708eV, 0.754eV and 0.778eV for different fading times 5s, 10s, 100s, 200s, 300s, 600s, 1200s, 2400s, 17h and 100h respectively. Matsuzawa, T. et al. suggested that a trap depth around 0.65eV is optimal for a LLP [12]. All fading processes were performed in a darkroom at room temperature after 5min’s irradiation of a set of 254nm (8W) + 365nm (8W) ultraviolet lamps. All the measures were taken on the same portion of the phosphor in the amount of 1.7mg. The heating rates of all the thermoluminescence measures were 1K/s. It is worth mentioning that though a series of separated energy levels were received, they still should be as a whole considered a description of a continuous distribution rather than a list of independent trap depths, as there is also other evidence for the continuous distribution condition.

 

Fig. 3 Fading thermoluminescence curves (FTLCs) which were measured at different fading times 5s, 10s, 100s, 200s, 300s, 600s, 1200s, 2400s, 17h and 100h, respectively. Illumination was performed by a set of 254nm (8W) + 365nm (8W) ultraviolet lamps. Illumination time was 5min. Thermoluminescence was measured at a heating rate of 1K/s and 1.7mg sample was used.

Download Full Size | PDF

All FTLGs in Fig. 3 had their high-temperature sides overlapped closely and the fading process traps were released from lower to higher temperature traps sequentially. Noticeably, even the 100h curve did not show obvious release of its high-temperature traps. This phenomenon is different from that which Joanna Trojan-Piegza et al. reported in LLP Lu₂O₃:Tb³⁺, Ca²⁺ which had separated trap depths [33]. In their work, TL curves with different fading times 3min, 30min, 60min and 120min did not overlap their high-temperature sides, and had a tendency to shrink to the peak’s position, which presented one of the separated trap depths from both low and high-temperature sides. This tendency was very obvious when 3min curve and 120min curve were compared.

The broadness of the TL peaks shown in Fig. 3 also suggested a continuous distribution [21, 22, 24]. Shown in Fig. 3, there was also a slightly upper convex section around 150°C in which all the FTLCs bent slightly, especially for the most faded ones 2400s, 17h and 100h curves. It is possible that another cluster of continuous distribution locates in this section. In that case, valleys in Fig. 5 can be better signs in distinguishing these trap clusters.

Figure 4 shows the same data as in Fig. 3 but in an Arrhenius diagram in order to evaluate trap depths of the FTLGs through the initial rise method. The slopes of the straight parts on the low-temperature sides give an expression $-slop=E_T/k_B$ about the trap depths when units are eV and K. Comparing with the Urbach's peak position method (shown in Table 1), the initial rise method gave a deeper and wider trap depths distribution. Trap depths of the FTLGs in Fig. 4 are 0.441eV, 0.491eV, 0.669eV, 0.760eV, 0.766eV, 0.878eV, 0.882eV, 0.975eV, 0.984eV and 1.074eV for different fading times 5s, 10s, 100s, 200s, 300s, 600s, 1200s, 2400s, 17h and 100h, respectively. Despite quantitative differences, these two methods both suggested that curves with different fading times represented different trap depths. This means that to find a single characteristic temperature of this kind of phosphors by simply using concept of trap depth would hardly work and other characterization quantities were needed.

 

Fig. 4 Fading thermoluminescence curves (FTLCs) in an Arrhenius diagram. When energy unit is eV and temperature unit is K, the slopes of the straight parts on the low-temperature sides give an expression $-slop=E_T/k_B$which can be used in evaluating trap depths of the FTLGs. Dash line marks an approximate boundary of data scope which can be used in the initial rise method (the straight line portion).

Download Full Size | PDF

Table 1. Trap depths estimated by Urbach's peak position method and the initial rise method. Unit is eV.

View Table

In Fig. 5, 5s the curve was the same as in Fig. 3 but all other curves were obtained by subtracting FTLGs from the 5s curve separately. The curves (except the 5s curve) in this figure show traps released in time windows from 5s to their fading times 10s, 100s, 200s, 300s, 600s, 1200s, 2400s, 17h and 100h, respectively. There was also an overlap but in low-temperature sides in Fig. 5 which means releases of traps processed from low temperature to high temperature sequentially. All the obtained curves show very low intensity around 150°C in Fig. 5(a) which is a consequence of high-temperature sides overlapping in Fig. 3. As noted above, this one-side shrinking of FTLGs suggested there was no center, namely a determinate separated trap depth, to shrink to in the fading process. As shown in Fig. 5(a), there were valleys around 150°C. For the afterglow decay of SrSiAl₂N₂O₃:0.001Eu²⁺ LPP has already reached 0.32mcd/m² at 2490s, it is safe to say that in the time window 5s~100h all traps related to persistence luminescence at room temperature were released. Thus the valley of the 100h curve around 150°C would give a reliable estimate of maximum trap depth at which electrons or holes can escape by thermal stimulation at room temperature. In detail, Fig. 5(b) shows valleys around 150°C were located between 140°C and 150°C except the 100s curve which was located around 100°C. As Urbach's method only needs a temperature to give an energy value, we used this method and got an energy range of 0.826~0.846eV which was 27%~30% higher than the optimal depth 0.65eV for a LLP [12]. On the one hand, even if these valleys have origins related to another cluster of continuous distribution around 150°C as what mentioned before, the same one-sides shrinking behavior of FTLGs still suggests traps at this or deeper depth regions are hardly to take part in the afterglow process.

 

Fig. 5 All curves obtained by subtracting fading thermoluminescence curves (FTLCs) from the 5s curve separately, except the 5s curve which was the same curve in Fig. 3 and (B) shows details of (A) around 150°C. The arrow pointed at the valleys.

Download Full Size | PDF

4. Conclusion

In summary, we reported persistent luminescence and thermoluminescence property of a bluish-white long persistence phosphor SrSiAl₂N₂O₃:0.001Eu²⁺. It had an afterglow persistent time of about 2490s and a continuous distribution of trap depths. This continuity of trap depths distribution was recognized from its $t^{-1}$ decaying behavior of afterglow [21, 22], highly high-temperature sides overlapping of FTLGs and broadness of the TL peaks [21, 22, 24]. We also determined the maximum trap depth of thermal stimulation at room temperature, which is the upper limit of trap depths having contribution to the afterglow performance. Indeed, this experimental value was obtained by using continuity of the phosphor’s trap depths distribution, but it also can be a reference value for LPPs without this kind of continuity.

As the special thermoluminescence theories and persistent luminescence theories for the continuous distributed condition are still at an exploration stage, the experimental studies of this kind of LPPs, such as determining and characterizing the nature of the traps, are necessary to establish bases of these special theories.