1. Introduction

Long persistent luminescence is a luminescent phenomenon that can last for minutes to hours, usually at room temperature (RT), after ceasing excitation sources. This process can be qualitatively explained by an electron transfer assumption [1, 2]: when persistent phosphors are excited by ultraviolet (UV) light (visible light in some cases), electron-hole (e-h) pairs are generated and the excited electrons from the conduction band (CB) are captured by electron traps. This process is usually called trapping process. Then the stored electrons can be liberated by thermal stimulations to CB (detrapping process) being recaptured by a luminescent center, resulting in a kind of thermoluminescence (TL). Thus, the key point to design persistent phosphors with long duration is the possibility of tuning the “trap depth” between the bottom of CB and the electron trap to form a proper energy gap favorable for the detrapping process [1–4]. If the binding energy of an electron trap is predicted, this strategy to control the trap depth by changing the CB energy level is regarded as “CB engineering”.

Recently, the CB engineering has been applied and proved to be a highly desirable strategy to develop new persistent phosphors in some aluminate/gallate based garnets (i.e. Ce³⁺-doped YSGG (green) [5], YAGG (green) [6] and GAGG (yellow) [7] phosphors), in which the possibility of photoionization from the excited 5d₁ energy level of Ce³⁺ to the bottom of CB with blue illumination can be increased by increasing Ga³⁺ substitution content, x for Al³⁺ [8, 9]. We have also developed red persistent phosphor (~690 nm), Cr³⁺ doped YAGG with similar garnet matrix, where Cr³⁺ ions act both as emission centers and trap centers [10]. The persistent behavior of this material can systematically be tuned by changing Ga³⁺ content. The persistent radiance (in unit of mW/Sr/m²) of the optimized composition was even higher than that of the ZnGa₂O₄:Cr³⁺ phosphor, which is used most widely for in vivo bio-imaging applications [11–14] since it was firstly reported by a French group [15]. Hence, the successes of CB engineering by Ga³⁺ substitution in garnets encourage us to design improved red persistent phosphors with brighter luminescence and longer afterglow.

The design concept in this paper is based on a theoretical assumption according to the Dorenbos theory [16, 17] for garnet compounds [18]. This diagram proposed by Dorenbos provides strong predicting power since the characteristic variation in electron and hole trapping depths of lanthanide ions is given by the shape of the two zigzag curves representing the ground state (GS) of divalent and trivalent lanthanide ions, both of which show a minimum bottom at 4f⁷ configuration (Eu²⁺ and Gd³⁺) [16]. Because the zigzag shape of two curves remain nearly unchanged in different matrices, once the binding energy of the GS for one lanthanide ion relative to the host CB or valence band (VB) is determined, those of 4f (sometimes 5d) levels of all other 13 lanthanides can be estimated fairly well by the Dorenbos diagram. The suitability of Dy³⁺ and Nd³⁺ ions as excellent electron traps in Eu²⁺-doped SrAl₂O₄ and CaAl₂O₄ persistent phosphors, respectively, was also explained well [17], and the optimal trap depth for the longest persistent phosphors, SrAl₂O₄:Eu²⁺-Dy³⁺ working at RT was proved to be ~0.65 eV [19]. Among all the trivalent lanthanide ions, Eu³⁺ ion would become the deepest electron trap to be Eu²⁺ (4f⁷) and thus is considered to hardly act as a good sensitizer in most persistent phosphors working at RT. According to former results reported by Dorenbos’s group, the energy level of Eu²⁺ GS is about 1.62 eV below CB in Y₃Al₅O₁₂ (YAG) [20]. Since the band-gap energy of YAG is 7.67 eV and Gd₃Al_5-xGa_xO₁₂ is 6.80 eV (x = 3), 6.53 eV (x = 4) and 6.48 eV (x = 5), respectively [18], considering the appropriate energy difference between YAG and Gd₃Al_5-xGa_xO₁₂, we assume that the energy level of Eu²⁺ in GAGG and/or GGG may be suitable for persistent luminescence.

Based on these ideas, we choose Eu³⁺ as a potential sensitizer in the Cr³⁺-doped Gd₃Al_5-xGa_xO₁₂ (GAGG, x = 3~5) composition [21] to design novel deep-red persistent phosphors with longer wavelength due to its weak crystal field. We found that the highest Ga³⁺ substitution (x = 5) gave the best persistent luminescent properties among GAGG:Cr³⁺-Eu³⁺ phosphors.

2. Experimental

Transparent ceramics (TCs) of GAGG:Cr³⁺-Eu³⁺ and GGG:Cr³⁺ with composition of (Eu_0.007Gd_0.993)₃Al_5-xGa_xO₁₂ (x = 3, 4, 5) and Gd₃Ga₅O₁₂ doped with 0.05 mol% Cr³⁺ were fabricated by one step solid-state reaction method using vacuum sintering. The detailed information of preparation procedures is similar to our previous reports [7, 22].

The in-line optical transmittance of GAGG:Cr³⁺-Eu³⁺ TCs was measured by UV-VIS-NIR spectrometer (Shimadzu, UV-3600). Microstructure observations including surface and fractured surface were examined by SEM (JEOL, JSM-890). Photoluminescence (PL) and persistent luminescence (PersL) spectra of the ceramic sample were measured by a CCD spectrometer (Ocean Optics, QE65-Pro) connected with an optical fiber. A combination of a 300 W Xe lamp (Asahi Spectra, MAX-302) with a UV mirror module (250-380 nm) was used as the excitation source for TL measurements. The ceramic sample was set in a cryostat (Advanced Research Systems, Helitran LT3) to control temperatures and firstly excited by UV light at 130 K for 10 min, then heated up to 600 K at a rate of 10 K/min after ceasing illumination for 10 min. The same CCD spectrometer was operated simultaneously with the TL measurement to always monitor TL spectra of the sample at different temperatures. Persistent luminescent decay curves were measured at 25°C using the same photodiode as the TL measurement (Electro-Optical Systems, S-025-H). Then, the decay curves were calibrated to the absolute radiance (in unit of mW/Sr/m²) using a radiance meter (B&W Tek Inc, Glacier X). Photographs of the TCs were taken by a digital camera (EOS kiss X5, Canon).

3. Results and discussion

Figure 1(a) shows the PL spectrum of the GAGG:Cr³⁺-Eu³⁺ (x = 3) sample under UV (254 nm) excitation. The sample shows sharp luminescence bands owing to the 4f-4f transitions of Eu³⁺: ⁵D₀→⁷F_J (J = 0, 1, 2, 3, 4, 5, 6). The intense bands observed are due to ⁵D₀→⁷F₄ at 708 nm and ⁵D₀→⁷F₁ (magnetic dipole, MD) transition at 590 nm. Generally, the probability of ⁵D₀→⁷F₂ electric dipole (ED) transition at 608~618 nm is very sensitive to the asymmetry of Eu³⁺ site [23], while that of MD transition is independent of it. The ratio of the integrated intensity of the ⁵D₀→⁷F₂ to ⁵D₀→⁷F₁ transitions is much smaller compared with Y₂O₃ [24] and Y₂O₂S [25], where Eu³⁺ ions take the Y-site with very low symmetry, resulting in strong luminescence at 610~620 nm with better color coordinates as a red phosphor. This small ratio indicates that Eu³⁺ ions substitute Gd³⁺ ions at dodecahedral sites in garnet, which is coordinated by eight O^2- ions in D₂ point symmetry.

 

Fig. 1 (a) PL and (b) PersL spectra of the GAGG:Cr³⁺-Eu³⁺ transparent ceramic (x = 3).

Download Full Size | PDF

After ceasing the UV excitation, the typical Eu³⁺ sharp emissions totally disappeared, instead, intra-3d transitions of Cr³⁺ composed of sharp R-line: ²E→⁴A₂ peaked at 695 nm and a broad emission band due to ⁴T₂→⁴A₂ peaked at 713 nm become dominant as shown in Fig. 1(b). The complete difference between PL and PersL spectra of the GAGG:Cr³⁺-Eu³⁺ phosphor shows that the emission centers change from Eu³⁺ to Cr³⁺ ions after stopping the UV illumination. Since the 10Dq value of Cr³⁺ ion in GAGG with larger x (Ga content) is expected to be lower than that in YAG or GAGG with smaller x, the broad ⁴T₂ band at longer wavelength becomes more dominant than the R-line from ²E level, which is dominant in a host with a large 10Dq such as ruby [26]. This situation makes the mean wavelength of PersL in GAGG longer than those in Cr³⁺ singly-doped YAGG with the same or smaller x [10].

Figure 2(a) shows the photograph of the GAGG:Cr³⁺-Eu³⁺ TCs (x = 3, 4, 5, thickness of 1 mm), through which we can clearly recognize the words below by naked eye owing to their high optical transparency. Two broad absorption bands are observed centered at about 620 and 450 nm due to transitions of Cr³⁺: ⁴A₂(⁴F)→⁴T₂(⁴F) and ⁴A₂(⁴F)→⁴T₁(⁴F), respectively. The weakness of the two bands is mainly due to the low doping concentration of Cr³⁺ (0.05 mol%). Besides, three sharp absorptions centered at 395, 312, 275 nm are ascribed to typical f-f transitions of Eu³⁺: ⁷F₀→⁵L₆, ⁷F₀→⁵H₃, ⁷F₀→⁵F₂, respectively and one intense absorption below 250 nm is attributed to the charge transfer band (CTB) of Eu³⁺-O^2-. Figures 2(b)-2(c) give the SEM observations of the polished surface and fractured surface of the GAGG:Cr³⁺-Eu³⁺ (x = 3) ceramic sample. Almost no micro-pores or secondary phases are observed either on the surface, inside of grains or at grain boundaries leading to its high optical transparency.

 

Fig. 2 (a) In-line optical transmittance and photograph of GAGG:Cr³⁺-Eu³⁺ transparent ceramics (thickness of 1 mm) with different Ga³⁺ contents as well as SEM observations of the (b) polished surface (c) fractured surface of the GAGG:Cr³⁺-Eu³⁺ (x = 3) transparent ceramic.

Download Full Size | PDF

Figures 3(a)-3(c) show the photographs of the three bulk samples under and after UV (254 nm) illumination. Under UV excitation, all of the TCs look orange mainly due to the 590 nm PL band since the photopic sensitivity of human eyes becomes steeply higher in the wavelength region less than 620 nm [4], while they look deep-red after ceasing the excitation. These color observations accord with the spectra shown in Fig. 1(a) and 1(b).

 

Fig. 3 Photographs of the GAGG:Cr³⁺-Eu³⁺ transparent ceramics (thickness of 1 mm) with different Ga³⁺ contents (x = 3, 4, 5) (a) under UV (254 nm) lamp (exposure of camera: 0.05 s) and (b) 30 s (c) 60 s after ceasing UV illumination (exposure of camera: 10 s), respectively. (d) persistent decay curves of the GAGG:Cr³⁺-Eu³⁺ transparent ceramic phosphors with different Ga³⁺ contents

Download Full Size | PDF

Persistent luminescent decay curves of the GAGG:Cr³⁺-Eu³⁺ ceramic phosphors (x = 3, 4, 5) after ceasing 5 min UV illumination are shown in Fig. 3(d). The persistent decay curves of the standard ZnGa₂O₄:Cr³⁺ ceramic phosphor under the same experimental condition previously reported by our group [27] and a Cr³⁺ singly doped GGG ceramic phosphor (x = 5) are also plotted as references. The radiance values at 5 min, 30 min and 60 min after ceasing the excitation of the samples are summarized in Table 1.The persistent radiance of GAGG:Cr³⁺-Eu³⁺ phosphors increases monotonously with the increase of Ga³⁺ content, x. Furthermore, the persistent radiance at 1 h of the GGG:Cr³⁺-Eu³⁺ (x = 5) sample (0.97 × 10⁻¹ mW/Sr/m²) is approximately 25 times higher than that of the Cr³⁺ singly doped GGG sample (0.04 × 10⁻¹ mW/Sr/m²), and is over 6 times higher than that of the widely used ZnGa₂O₄:Cr³⁺ red persistent phosphor (0.15 × 10⁻¹ mW/Sr/m²) [27].

Table 1. Radiance of GAGG:Cr³⁺-Eu³⁺ transparent ceramic phosphors with different Ga³⁺ contents compared with GGG:Cr³⁺ and ZnGa₂O₄:Cr³⁺ ceramic phosphors [27].

View Table

TL glow curves of the GAGG:Cr³⁺-Eu³⁺ samples (x = 3~5) are shown in Fig. 4(a).The TL glow peak shifts monotonously from higher to lower temperature with increasing Ga³⁺ content (498 K, 408 K and 355 K for x = 3, x = 4, x = 5, respectively). Since the TL peak temperature is correlated to the trap depth [1], the downshift of the glow peak indicates that the trap depth becomes shallower. In Fig. 4(b), two-dimensional (2D) mapping of TL intensity was plotted in order to clarify what kind of emission contributes to the TL glow peak at different temperatures in the GGG:Cr³⁺-Eu³⁺ phosphor (x = 5, showing highest persistent brightness). From the contour mapping, it can be seen that at increased temperatures, the TL glow peak is simply composed of broad emission bands due to Cr³⁺ centered at about 730 nm. No sharp f-f transitions of Eu³⁺ are observed at any temperatures. Therefore, we conclude that the Eu³⁺ acts as an electron trapping center and changes into Eu²⁺ (Eu³⁺-e^-) being located below CB. A schematic model is described by the inserted schematic illustration in Fig. 4(a). The detrapping rate, which is dominated by the trap depth (as well as temperature and frequency factor) [1], becomes faster with increasing Ga³⁺ content due to the downshift of CB level. Generally, not only in conventional phosphors for illumination applications, europium is also widely used as emission centers in persistent phosphors like well-known SrAl₂O₄:Eu²⁺-Dy³⁺ (green) [19], CaAl₂O₄:Eu²⁺-Nd³⁺ (blue) [28] and Y₂O₂S:Eu³⁺-Ti⁴⁺-Mg²⁺ (red) [29] phosphors. In this study, we lowered the CB level to decrease the trap depth from Eu²⁺ GS as possible as we can by full Ga³⁺ substitution for Al³⁺ in GAGG to finally be GGG, and have successfully developed a novel bright deep-red Cr³⁺ doped persistent phosphor in which Eu³⁺ ions act as an electron trap. A phosphor with such an unusual combination of emission and trap centers has never been reported before and we thus state as a new discovery.

 

Fig. 4 (a) Thermoluminescence (TL) glow curves of the GAGG:Cr³⁺-Eu³⁺ transparent ceramic phosphors with different Ga³⁺ contents, schematic illustration inserted (b) wavelength-temperature (λ-T) contour plot of the GGG:Cr³⁺-Eu³⁺ (x = 5) transparent ceramic.

Download Full Size | PDF

4. Conclusions

In summary, we successfully developed a series of transparent ceramic phosphors: GAGG:Cr³⁺-Eu³⁺ with deep-red persistent luminescence and long duration time, in which Eu³⁺ ions act as an electron trap. The persistent radiance of the GGG:Cr³⁺-Eu³⁺ (x = 5) sample is over 6 times higher than that of the widely used ZnGa₂O₄:Cr³⁺ red persistent phosphor and its broad band persistent emission centered at 730 nm possesses deeper penetration capability through biological tissues. Practical applications in the in vivo bio-imaging field can be expected in the near future by using GGG:Cr³⁺-Eu³⁺ nano-particles as optical probes.