1. Introduction

Long persistent luminescence (LPL) is a phenomenon that the photoluminescence of materials doesn’t disappear for an obvious length of time when the excitation source is turned off [1–3]. The LPL phosphors exhibit particular light storing and releasing properties. In recent years, researchers have paid considerable attention on the investigation of LPL phosphors because this type of material is reusable and energy-storable, and now has been extensively applied in emergency signing, luminous paints, escape route indicators, decorative purposes and watch dials, even in medical inspection [4–7]. Up to now, the best performance of LPL phosphors has been reported as Eu²⁺-doped alkaline earth aluminates or silicates, which emit in the green or blue region [8,9]. Eu²⁺, Dy³⁺ co-doped SrAl₂O₄ phosphor and its derivatives are well known to their excellent characteristics in terms of high luminance, long lasting time and high stability compared to previous sulfide phosphors. However, the synthetic temperature of these materials is usually too high (>1200°C), which disadvantages the development of the application for the expensive cost. On the other hand, the emitting centers in LPL phosphors have been focused on the discrete luminescent centers, e.g. rare earth ions [10,11]. However, except the discrete luminescent centers, the excition centers derived from the introduction of trapping levels are also significant role in the transition process. Unfortunately, details information such as the nature and origin of the defects in LPL are still a subject of challenge, and the transport process of the charge carriers between trap and emission center remains unclear. Therefore, simplifying the transport process of the carriers or exploring the trapping centers acting as activators simultaneous, is benefit us to further understand the trapping mechanism in LPL.

At present, germanates have been developed as LPL phosphors due to their low synthesis temperature, high stability and reasonable conductivity as an oxide host [12,13]. In this work, a unique self-activated LPL phenomenon in Ca₂Ge₇O₁₆ phosphor is observed for the first time. Furthermore, both the PL and LPL peaks in Ca₂Ge₇O₁₆ phosphors shift to long wavelength obviously with increasing concentration of Nd³⁺, which contributes to the corresponding emitting color changing from purple to blue.

2. Experimental

Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.03, 0.05, 0.07) samples were synthesized by the conventional high temperature solid state reaction. The stoichiometry amounts of CaCO₃(A.R), GeO₂(A.R), Nd₂O₃(99.99%) were mixed in an agate mortar. After fully grinding, the mixtures were put into crucibles and calcined at 1000 °C for 12 h in air and 10⁻² Torr vacuum atmosphere. After annealing, the samples were cooled to room temperature in the furnace.

The crystalline structures of the prepared powders were investigated by X-ray diffraction (XRD) with Ni-filter Cu Kα radiation(λ = 0.154056 nm) at a scanning stepping of 0.02°. The XRD data were collected in the range of 10° to 70° by applying a D8ADVANCE/Germany Bruker X-ray diffractometer. The absorption spectra were obtained using the UV-visible spectrophotometer. The photoluminescence excitation (PLE), emission (PL) spectra and LPL were recorded by using a Hitachi F-7000 fluorescence spectrophotometer. LPL decay curves were measured with a PR305 long afterglow instrument after the sample irradiated by 254 nm for about 10 min. The thermoluminescence (TL) curves were measured with a FJ-427 A TL meter (Beijing Nuclear Instrument Factory). The samples weight was kept constant (0.002g). Prior to the TL measurement, the samples were first exposed to the radiation of 254 nm for about 10 min, then heated from room temperature to 500K with a rate of 1K/s.

3. Results and discussion

The crystal phase was characterized by XRD measurements. Figure 1(a) shows the Rietveld structural refinements of the powder diffraction patterns of Ca_1.99Ge₇O₁₆: 0.01Nd³⁺ phosphor. The black solid lines and red crosses are calculated patterns and experimental patterns, respectively. The black short vertical lines show the position of the Bragg reflections of the calculated pattern. The difference between the experimental and calculated patterns is plotted by the green line at the bottom. The reliability parameters of refinement are R_wp = 12.4%, and χ² = 1.27, which can verifies the phase purity of the as-prepared samples. Figures 1(b)-1(c) show the X-ray diffraction patterns of Ca₂Ge₇O₁₆, Ca_1.95Ge₇O₁₆: 0.05Nd³⁺ and the JCPDS Standard Card (No. 34-0286) of Ca₂Ge₇O₁₆, respectively. All the diffraction peaks can be indexed to Ca₂Ge₇O₁₆ phase. No impurity phase is observed in these samples, demonstrating the pure phase was synthesized in this work. Based on the effective ionic radius (r) of cations with different coordination number (CN), the radii of Nd³⁺ (0.0995nm) is closer to that of Ca²⁺ (0.0990nm). Therefore, Nd³⁺ substitutes the Ca²⁺ sites in Ca₂Ge₇O₁₆ host. The lattice constants are a = b = 11.340(2)Å, c = 4.6400Å and V = 596.68(23) ų, Z = 2. Figure 1(d) represents the structure diagram of the Ca₂Ge₇O₁₆ host, which shows that Ca ions are located at the 4g site, Ge ions are located at the 8a, 2d and 4g site, and O ions are located at the 8i site. Structurally, Ca₂Ge₇O₁₆ belongs to the orthorhombic with the space group Pb_a/2 (32). A sheet of four-membered rings of Ge tetrahedra (with Ge on the 8i position) and isolated Ge tetrahedra (Ge on the 4g position) alternate with a sheet of Ge octahedra (Ge on the 2d position) and eightfold-coordinated Ca sites along the c direction in an ABABA... Sequence [14].

 

Fig. 1 The experiment, calculated of the XRD refinement of Ca_1.99Ge₇O₁₆: 0.01Nd³⁺ (a); XRD patterns of the air-sintered Ca_1.95Ge₇O₁₆: 0.05Nd³⁺ (b); The vacuum-sintered Ca₂Ge₇O₁₆ and JCPDS Standard Card (No. 34-0286) of Ca₂Ge₇O₁₆ (c); The crystal structure of Ca₂Ge₇O₁₆ (d).

Download Full Size | PDF

The density functional theory calculations of Ca₂Ge₇O₁₆ based on crystal structure refinement were employed and shown in Figs. 2 (a)-(b). The local density approximation (LDA) was chosen for the theoretical basis of the density function. Ca₂Ge₇O₁₆ possessed a direct band-gap of about 3.518 eV with the valence band (VB) maximum and the conduction band (CB) minimum located at the G point of the Brillouin zone. It is expected that the value of the calculated band-gap of Ca₂Ge₇O₁₆ will be smaller than the experimental one as the LDA underestimates the size of the band-gap [13]. The inset of Fig. 2(b) plots the experimental absorption spectrum of Ca₂Ge₇O₁₆ crystal, which indicates that the host lattice absorption located at about 298 nm (4.161 eV).

 

Fig. 2 Energy band structure of Ca₂Ge₇O₁₆ crystal (a); Total density of states of Ca₂Ge₇O₁₆, the inset shows the absorption curve of Ca₂Ge₇O₁₆ (b).

Download Full Size | PDF

Figure 3(a) shows the PLE and PL spectra of the vacuum-sintered and air-sintered Ca₂Ge₇O₁₆ phosphors, respectively. Under the excitation at 245 nm, both the vacuum and air sintered Ca₂Ge₇O₁₆ host matrix exhibit a self-activated emission with a purple broad band ranging from 310 to 600 nm (centered at ~400 nm). The PLE spectra show a broad excitation band with the maximum at 245 nm (monitored at 400 nm). The energy of effective excitation peak (locates at 245 nm) is higher than that of the band gap calculated above, therefore, the electrons in the VB can be excited to the CB under 245 nm excitation. Compared with the air-sintered phosphor (blue curve), the emission intensity of vacuum-sintered sample increases about 300% (red curve). It indicates that sintering in vacuum is quite effective to improve the intensity of the self-activated luminescence in Ca₂Ge₇O₁₆. It is expected that abundant $V_O^{••}$ could be created when samples sintering under an oxygen-deficient (vacuum) atmosphere as proposed by many other groups [15–17]. Therefore, it could be conclude that the emission peak (~400 nm) is mainly associated with $V_O^{••}$, and the increasing intensity of the vacuum sintered sample was ascribed to the rise of total number of the exciton energy-level derived from the abundant $V_O^{••}$. An interesting result of this work is that the LPL phenomenon is observed in Ca₂Ge₇O₁₆ host matrix after the removal of UV excitation, as shown in Fig. 3(b). It could be noticed that the LPL spectra of the vacuum and air sintered Ca₂Ge₇O₁₆ exhibit similar shape and position to the PL spectra, indicating that the PL and LPL originate from the same emitting centers. Furthermore, the LPL intensity of the vacuum sintered Ca₂Ge₇O₁₆ phosphor (red curve) is much stronger than that of the air sintered one (blue curve), implying that the LPL properties are associated with $V_O^{••}$ as well.

 

Fig. 3 PLE (λ_em = 400) and PL (λ_ex = 245) spectra of the vacuum and air sintered Ca₂Ge₇O₁₆ phosphors(a); LPL spectra of the vacuum and air sintered Ca₂Ge₇O₁₆ phosphors (b).

Download Full Size | PDF

The spectra of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) under the excitation at 245 nm are presented in Fig. 4. A broad emission band located at in the range of 310-600 nm is observed in Nd³⁺ doped samples. Although the characteristic emission of Nd³⁺ is not detected in visible light, the red-shift and the asymmetry of the emission peaks are demonstrated in these samples with increasing concentration of Nd³⁺, which contributes to the corresponding emitting color changes from purple to blue. The emission intensity enhances gradually with increasing concentration of Nd³⁺, and reaches its maximum when x = 0.005. In addition, accompanying the variation of emission intensity, an obvious red-shift of emission peaks is observed. The inset shows the Gaussian profiles of Ca_1.95Ge₇O₁₆: 0.05Nd³⁺, the PL spectrum can be roughly fitted into two peaks centered at 400 nm (peak A) and 476 nm (peak B). In consideration of the fact that Nd³⁺ ion replaces the Ca²⁺ site in Ca₂Ge₇O₁₆ host matrix, two Nd³⁺ ions replace three Ca²⁺ ions to balance the charge of the phosphor, which create two positive defects and one negative defect. As the hole traps of negative defect, Ca vacancies ($V_{Ca}^{\text{'}\text{'}}$) are formed by $2N{d}^{3+}+3C{a}^{2+}\to 2N{d}_{Ca}^{•}+V_{Ca}^{\text{'}\text{'}}$. So, the formation of the charge compensating lattice defects of $V_{Ca}^{\text{'}\text{'}}$ is unavoidable. It could be expected that the number of $V_{Ca}^{\text{'}\text{'}}$ increases with increasing concentration of Nd³⁺, which contributes to the higher intensity of peak B. Meanwhile, the ionization potential of Nd³⁺ appears lower than that of Ca²⁺ ions, so the Nd³⁺ ion has greater ability to stabilize $V_O^{••}$ [18,19]. Therefore, the number of the $V_O^{••}$ increases with increasing Nd³⁺ concentration as well, which results in the higher intensity of peak A. Thus, it could be concluded that the stronger emission intensity of Nd³⁺ doped Ca₂Ge₇O₁₆ phosphors derives from the increased number of $V_O^{••}$ and $V_{Ca}^{\text{'}\text{'}}$. Moreover, the number of $V_{Ca}^{\text{'}\text{'}}$ increases linearly with increasing concentration of Nd³⁺, while the intrinsic defects of $V_O^{••}$ reach the saturation after the gradually initial increase. Therefore, the differential increase rate of $V_O^{••}$ and $V_{Ca}^{\text{'}\text{'}}$, that the increasing rate of $V_{Ca}^{\text{'}\text{'}}$ is higher than that of $V_O^{••}$, or the decreasing rate of $V_{Ca}^{\text{'}\text{'}}$ is slower than that of $V_O^{••}$, may contributes to the red-shift of the PL peak.

 

Fig. 4 PL spectra of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07), the inset shows the Gaussian profiles of Ca_1.95Ge₇O₁₆: 0.05Nd³⁺.

Download Full Size | PDF

To further understand the PL properties, the PL spectra of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphor irradiated with different time under the excitation at 245 nm are recorded in Fig. 5. Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ exhibits relative lower emission intensity initially, then gradually increases with prolonging irradiated time and reaches the maximum after 120 s. It can be speculated that there exists some traps in Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphors, which capture parts of excited carriers. It takes a certain amount of time to fill the traps saturated, which lead to the PL intensity weak at the initial time. In other words, it needs a certain amount of time to achieve the intrinsic emission intensity of Ca₂Ge₇O₁₆: Nd³⁺ after the capturing processes of carriers completed as described above.

 

Fig. 5 PL spectra of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphor with different irradiated time under the excitation at 254 nm.

Download Full Size | PDF

Figure 6 reveals the LPL spectra of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) measured 5 min after the removal of the excitation source. The LPL spectrum of Ca₂Ge₇O₁₆ phosphor with the absence of Nd³⁺ exhibits a symmetrical curve at 400 nm, while asymmetric LPL peaks was observed in Nd³⁺ doped samples. The inset of Fig. 6 shows the Gaussian profiles of Ca_1.999Ge₇O₁₆: 0.001Nd³⁺ in the range from 310 to 600 nm, which can be well-decomposed into two Gaussian components at 400 and 476 nm. The LPL intensity increases with increasing concentration of Nd³⁺ initial, of which reaches the maximum intensity for x = 0.005, and then decreases with further increase concentration of Nd³⁺. Moreover, the dominant position of LPL peaks gradually shift from 400 to 476 nm, which imply that $V_{Ca}^{\text{'}\text{'}}$ acting as the trap centers also plays a significant role on the LPL peaked at 476 nm.

 

Fig. 6 The LPL spectra of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) measured 5 min after removal of the excitation source.

Download Full Size | PDF

The decay curves of LPL of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) are measured at room temperature and displayed in Fig. 7. All doped samples show LPL with different persistent times, and consist of a fast decay and a consequent slow decay with a long decay tail, implying the existence of various trap depths [20]. It is known that the lattice defects acting as traps play an essential role for energy storage in LPL phosphors. Compared with the relatively weak LPL of the un-doped sample, longer persistent time is observed in the phosphors doping with Nd³⁺. Hence, it could be safe to say that the incorporation of Nd³⁺ creates more defects (the intrinsic defects or foreign defects) in Ca₂Ge₇O₁₆ host lattice, which act as trapping centers and have a significant influence on the LPL performance. Comparatively speaking, the Nd³⁺ doped samples exhibit better LPL performance, indicating that the incorporation of Nd³⁺ creates lots of suitable traps benefited for LPL. The inset of Fig. 7 shows the CIE chromaticity coordinates profile of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) samples, and the chromaticity coordinates shift from (0.1521, 0.0344) to (0.1917, 0.2339) with increasing Nd³⁺ concentration, which indicates that the LPL color changes from purple to blue accordingly.

 

Fig. 7 LPL decay curves of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) phosphors, the inset is the CIE chromaticity coordinates profile of Ca_2-xGe₇O₁₆: xNd³⁺ samples.

Download Full Size | PDF

Figure 8 shows the LPL spectra of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphor at different times after the removal of the excitation source. The LPL peak of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphor shows an obviously red-shift from 447 to 470 nm as the delay time increase, which provides further evidence that there are two traps (T_A and T_B) contributing to the two different LPL processes, and the decay rate of the T_A (400 nm) is higher than that of T_B (476 nm) .

 

Fig. 8 LPL spectra of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ phosphor at different times after removal of the excitation source.

Download Full Size | PDF

In general, a suitable trap depth and high trap density plays a dominate role in the performance of LPL. TL measurement provides an efficient way to reveal the relevant information of traps. It is widely accepted that the trap density is approximately proportional to the TL intensity and the optimal TL peak is usually situated in the temperature range of 320-390 K for the excellent LPL properties. As shown in Fig. 9, The TL of the un-doped Ca₂Ge₇O₁₆ phosphor shows a symmetrical curve, and the TL peak located at 346 K (T_A) is related to the traps of the intrinsic defect $V_O^{••}$. In Nd³⁺-doped ones, two TL bands can be identified. Except for the intrinsic traps T_A, the deeper traps (T_B) have been generated. Such observations could be identified as a sign of the presence of at least two different traps in Ca₂Ge₇O₁₆: Nd³⁺. The TL curves of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺ can be generally fitted to two bands centered at 346 and 384 K. Based on the above analysis, it implies that $V_{Ca}^{\text{'}\text{'}}$ act as the trap centers is the significant role in TL band at 384 K. The TL intensity increases initially with increasing Nd³⁺ concentration, then reaches a maximum at x = 0.005, which was corresponding with the PL and LPL spectra. Besides, the TL bands shift to higher temperature band side with increasing Nd³⁺ concentration. In order to estimate the defect states in these samples, the classical fitting peak-shape methods developed by Chen et al are introduced [21]. And the trap depth of E is calculated from the glow-peak parameters by the following equation [22]:

 

Fig. 9 TL curves of Ca_2-xGe₇O₁₆: xNd³⁺ (x = 0, 0.001, 0.005, 0.01, 0.05, 0.07) phosphors after the UV irradiation. The inset shows the Gaussian profiles of Ca_1.995Ge₇O₁₆: 0.005Nd³⁺.

Download Full Size | PDF

In summary, there are two different defects of $V_O^{••}$and $V_{Ca}^{\text{'}\text{'}}$, act not only as the exciton energy-level attributed to the emission process, but also as the trapping centers attributed to the LPL process. The formation of $V_O^{••}$ is unavoidable for oxygen atoms escape from the crystal sites under the high temperature during the solid state preparation, which also have been reported as electron traps in many electron trapping materials phosphors [23–26]. Meanwhile, because of the replacement of Nd³⁺ with Ca²⁺ ions in Ca₂Ge₇O₁₆ host matrix, the natural consequence is the formation of the charge compensating lattice defects, e.g., $V_{Ca}^{\text{'}\text{'}}$, which act as hole traps captured the holes. Therefore, a mechanism account for the tunable LPL phenomena in Ca₂Ge₇O₁₆ phosphors with and without Nd³⁺ doped could be proposed, as schematically shown in Fig. 10. Under UV excitation, the electrons are first excited from the valance band (VB) ground state to the conduction band (CB) (step 1) in Ca₂Ge₇O₁₆ without Nd³⁺. Subsequently, the promoted electrons can be trapped by $V_O^{••}$(Trap 1) (step 2). Trap 1 could be filled after sufficient illumination time. Then, a part of electrons relax to Trap 1 by nonradiative relaxation from CB. Finally, the self-activated emission of Ca₂Ge₇O₁₆ arises from the recombination (step 3) of the excited state (Trap 1) and ground state (VB). After the stoppage of the irradiation, the electrons are released from the Trap 1 under the effect of the thermal radiation. The direct recombination of Trap 1 to VB (step 3) contributes to the LPL process. For Nd³⁺-doped ones, under the excitation at UV light, the step 1 and 2 happen. Meanwhile, the holes are created in VB, and then trapped by holes traps $V_{Ca}^{\text{'}\text{'}}$(Trap 2) (step 4). After plenty of time, Trap 1 and Trap 2 are filled. Then, a part of electrons from CB relax to Trap 1, and a portion of holes from VB are released to Trap 2 by the thermal radiation. Finally, the PL of Ca₂Ge₇O₁₆: Nd³⁺ appeared from the recombination (step 3 and 5) of the excited state (Trap 1) and ground state (Trap 2 and VB). After the removal of the UV light, the electrons and holes are released under the effect of the thermal radiation. The direct recombination of Trap 1 to VB (step 3) and Trap 1 to Trap 2 (step 5) dominates the LPL at 400 and 476 nm, respectively. The result indicates that the LPL process of step 3 is depended on the trap 1, while step 5 is primarily depended on both of the two traps (Trap 1 and Trap 2). In other word, the T_A is depended on Trap 1, and the T_B is rooted from the combined action result of Trap 1 and Trap 2, this is the reason why T_B is deeper than that of T_A.

 

Fig. 10 Possible schematic of the emission and LPL mechanism in Ca₂Ge₇O₁₆: Nd³⁺ phosphors.

Download Full Size | PDF

4. Conclusion

A self-activated LPL phosphor Ca₂Ge₇O₁₆, was synthesized by the high temperature solid state reaction. With the doping of Nd³⁺, the LPL intensity was enhanced greatly, which is attributed to the stability of the $V_O^{••}$ acted as electrons traps and the increase of the $V_{Ca}^{\text{'}\text{'}}$ acted as holes traps. The TL curves exhibit two different traps centers with the activation energy of ~0.5369 and ~0.6279 eV that play an essential role for LPL in Nd³⁺ doped samples. With increasing the concentration of Nd³⁺, the emission and LPL peaks of these phosphors red-shift obviously, and the corresponding color of PL and LPL changes from purple to blue. The results indicate that the two different traps act not only as the exciton energy level attributed to the emission process, but also as the trapping centers attributed to the LPL process. In our work, a novel idea is provided for LPL materials, it simplifies the transport process between trapping center and emitting center, which could benefit us to understand the trapping properties in LPL.