1. Introduction

Persistent luminescence is a phenomenon that phosphors can still emit for seconds, hours or longer even excitation stopped. Persistent luminescence is a thermoluminescence (TL) known as thermal stimulated luminescence (TSL) phenomenon at a specific temperature, e.g. at the room temperature [1]. Recently persistent phosphors have attracted more interests due to their good optical properties and wide applications such as night vision, temperature sensors, and medical diagnostics [2–4]. In the past two decades, efforts have been made to either design or achieve long duration phosphors. By now blue CaAl₂O₄:Eu²⁺,Nd³⁺ [5], green SrAl₂O₄:Eu²⁺,Dy³⁺ [6], and red sulfides as Y₂O₂S: Eu³⁺,Mg²⁺,Ti⁴⁺ [7] and CaS:Eu²⁺,Tm³⁺, Ce³⁺ [8] phosphors have been successfully commercialized. However, there exist challenges to develop red persistent emitters with better chemical stability and longer afterglow duration.

Zinc gallate is a self-activated host for reddish afterglow phosphors which emits blue light when excited by low voltage electrons or UV light [9] and it also gives rise to reddish emission when doped with transition metal ions, i.e. trivalent Cr³⁺ ions. It exhibits bright red long lasting luminescence at around 695 nm due to the ²E→⁴A₂ transition [10]. In general, when Ga³⁺ ions are substituted by the doped Cr³⁺, the B-sites under distorted octahedral coordination will be created to some degree and occupied by Cr³⁺ ions. The ⁴T₂ excited state of Cr³⁺ is therefore strongly influenced by the crystal field from the ligand anions, so that ZnGa₂O₄:Cr³⁺ phosphors emit the characteristic emission of Cr³⁺ ranged from visible to near infrared [11]. Recently the improvement and its enhancement mechanism of the persistent luminescence of ZnGa₂O₄:Cr³⁺ have been studied by co-doping Ge, Sn or Bi. Allix reported that there existed the anti-site coordination in real ZnGa₂O₄ crystal to a certain extent and it was resulted from the coexistence of tetrahedrally coordinated Ga³⁺ and octahedrally coordinated Zn²⁺. Improvement of Cr³⁺ persistent luminescence was directly due to the distorted octahedral sites from the substitution of two Ga³⁺ by one Zn²⁺ and one Ge⁴⁺ in Zn_1+xGa_2-2x(Ge/Sn) _xO₄ [12]. And Zhuang et al. also demonstrated that the anti-sites defects caused efficient traps favoring the Cr³⁺ long lasting luminescence with Bi³⁺-doping [13]. Huang et al. and Bessiere et al. have shown that a preset Zn deficiency benefits to the red long-lasting luminescence compare to the stoichiometric ZnGa₂O₄, revealing that the unoccupied interstices could admit the doped ions entering [14,15]. The above substitution and proposed mechanism gave the opportunity of selecting suitable dopant ions to develop phosphors with better red long-lasting brightness and time. Moreover, the partially inverted spinel structure could be inevitably formed when ZnGa₂O₄ was synthesized under high temperature or oxygen deficient conditions. The inversion defects or deficient A-sites were induced by the inverted spinel structure and bale to consequently improve the persistent performance of Cr³⁺-centers [9,15–17]. Meanwhile the density and depth of traps both play a key role to the afterglow properties in the long-lasting phosphors [18,19]. In this work, we try the co-doping of Si⁴⁺ to improve the persistent properties of ZnGa₂O₄: Cr³⁺ phosphor and better understand the mechanism of enhancement of persistent performance.

2.Experimental

The samples of ZnGa_1.995-xO₄: 0.5% Cr³⁺, xSi⁴⁺ (x = 0, 1mol%, 3mol% and 5mol%) were prepared by means of the high temperature solid-state reaction. Raw materials of ZnO powder (99.9%), Ga₂O₃ powder (99.9%), SiO₂ (99.99%) and Cr₂O₃ (99.9%) were taken in a stoichiometric molar ratio and all were mixed and ground thoroughly in an agate mortar. The mixture was then placed in a corundum crucible and sintered at 1300°C for 6 hours for sample without Si⁴⁺, and at 1200°C for 4 hours for the other three samples with Si⁴⁺. Similarly, the amorphous SiO₂ could be considered as both dopant of Si⁴⁺ and a flux in this experiment when compared with the effect of H₃BO₃ onto the formation of ZnGa₂O₄ in previous work [20]. Due to the existence of SiO₂, the sintering temperature of the spinel ZnGa₂O₄ crystal lowers and the particle size became larger.

The XRD patterns for the crystal phases of products were collected with the angle scanning from 10° to 70° (2θ) at a step of 0.02°, on a powder X-ray diffractometer (XRD, RigakuD/Max-3B) equipped with Cu-K_α source (λ = 1.5406Å). Scanning electron microscopy (SEM, Model No: JXM-6700F, Japan) was used to observe the morphology of the powders. The photo-excitation, luminescence (PL), and afterglow decay curves of samples were determined by a spectrometer (Edinburg FLS980) operated at room temperature. In which the afterglow decay curves were recorded after the samples were irradiated by 254 nm light for 5 min. A thermoluminescent dosimeter (model: FJ427A1, CNNC Beijing Nuclear Instrument Factory) was applied to measure the thermoluminescence (TL) glow curves. The TL curves were examined after the sample was irradiated by UV lamp for 5 min and placed in dark for 2 min and scanned from room temperature to 300 °C.

3.Results and analysis

3.1 XRD patterns

The X-ray diffraction patterns were presented in Fig. 1. As shown, all of the diffraction patterns are almost identical in both peak position and height with respect to the JCPDS Card 38-1240 (ZnGa₂O₄). There was no impurity phase to be observed in the XRD patterns of all samples. It was indicated that Cr³⁺ and Si⁴⁺ ions fully took part in the crystalline reaction and occupied the appropriate locations. The co-doping of Si⁴⁺ did not change structure of the host of ZnGa₂O₄.

 

Fig. 1 The XRD patterns of the prepared samples ZnGa₂O₄: 0.5% Cr³⁺, xSi⁴⁺ with (a) x = 0, (b) x = 0.01, (c) x = 0.03, (d) x = 0.05.

Download Full Size | PDF

The results were remarkably different from the results reported by Allix [12]. The GeO₂ phase was indeed visible in XRD patterns when they attempted to synthesize Zn_1+xGa_2-2x(Ge/Sn) _xO₄ with Ge or Sn doping. Considerably, the same tetravalent ions of Si⁴⁺ and Ge⁴⁺ might induce the similar results and then the Si⁴⁺ ions would mainly take the B-site coordination and small portion of A-site coordination. Unfortunately, no SiO₂ phase was observed even though more than 1mol% of Si⁴⁺ ions doped. The low sensitivity of XRD applied is one of the possible reasons to account for the observation. So the NMR or EDS measurements could supply more precise details about this substitution mechanism.

3.2 SEM images of phosphor powders

The luminescence property is strongly influenced by the morphology of phosphors. Figure 2 displayed the SEM images of the as-prepared samples. The results showed that the sintering temperature lowered along with the increasing SiO₂ concentration. The phosphor samples illustrated obvious structural agglomeration to some degree leading to bigger particle size and melting in shape. So the function of SiO₂ is a flux, and the addition of SiO₂ strongly affected the morphology and the persistent luminescence of phosphor samples. The further investigations are expected to obtain the more precise details about the effects of the amorphous SiO₂ and its substitution mechanism by relying on NMR or Raman detection.

 

Fig. 2 The SEM images of ZnGa₂O₄: 0.5% Cr³⁺, xSi⁴⁺ samples with (a) x = 0, (b) x = 0.01, (c) x = 0.03, and (d) x = 0.05.

Download Full Size | PDF

3.3 Photoluminescence properties

The photo-excitation spectra have been measured by monitoring λ = 687 nm emission as shown in Fig. 3. It exhibited three broad absorption bands within the range from 210 nm to 650 nm. In which the band centered at 260 nm was well known as the charge transfer band (CTB) corresponding to the transition of electron at O 2p orbital to 4s4p orbital of Ga. The other two bands centering at around 410 nm and 550 nm, could be assigned to ⁴A₂-⁴T₁ and ⁴A₂-⁴T₂ transition of the inter-3d electrons transitions of Cr³⁺, respectively [10,12,15]. It was found that the intensity of the 410nm became a bit stronger based on the resonant energy transfer from the host to Cr³⁺ ions. The resonant energy transfer could be judged from the overlapping between the host emission (ranged from ultraviolet to blue area) and the absorption band of Cr³⁺ (peaking at 410 nm), which allowing the doping Cr³⁺ to sequently absorb the energy from the emission of host leading to the enhanced ionic characteristic emission of Cr³⁺ [16].

 

Fig. 3 Excitation spectra of ZnGa₂O₄: 0.5% Cr³⁺ by monitoring at λ_em = 687 nm.

Download Full Size | PDF

The emission spectra of the phosphors samples under excitation at 400 nm are shown in Fig. 4, in which (a) presents the luminescence of ZnGa₂O₄: 0.5% Cr³⁺ without Si doping. It consists of series of narrow recognizable lines of Cr³⁺. The dominant 687 nm was unresolved zero phonon lines (R lines) known as R2 and R1 doublets from the splitting of ²E in a weak trigonal distortion field. Unfortunately they were unresolvable under the present resolution. The R lines are theoretically assigned to the emissions of Cr³⁺ taken the octahedrally coordinated “ideal” sites in spinel ZnGa₂O₄ [21]. Moreover, the phonon side bands (PSB) of the R lines were also clearly determined as 708 nm and 714 nm for the Stokes PSB, and 670 nm and 680 nm for the Anti-Stokes PSB, respectively. They were caused by the combination of the spin-orbit coupling and the trigonal field as marked in Fig. 4(a). The PSB positions and relative intensities keep in good accordance with those from infrared and Raman experiments [12,15,22]. Whilst the additional 695nm peak was labeled as N2 line and not attributed to the above mentioned mechanism. It has been considered as one of “structure dependent” transition from the other type of Cr³⁺ in contrast to the “ideal” coordination. And this kind of Cr³⁺ ions was affected by a perturbed short-range crystalline order differing from the ideal octahedral coordination in the normal spinel structure. This kind of Cr³⁺ was therefore located in a distorted environment affected by the antisite defect. As mentioned in literatures [12,13,15], the antisite defect was established by an inversion crystalline process and located in the first cationic neighbors of Cr³⁺. Given the R and N2 lines, at least two coordinated Cr³⁺ ions have been formed in either normal symmetrical or the inversion environment distorted by the antisite defect.

 

Fig. 4 Emission spectra of ZnGa2O4: 0.5% Cr3 + , xSi4 + by λex = 400 nm excitation.

Download Full Size | PDF

Figures 4(b)-4(d) exhibit the emission spectra of the other three phosphors doped with Si. It was worth noting that the intensity of N2 line relative to the R lines increases along with the increment of the contents of Si⁴⁺. Bessière et al explained that the change of the intensity of N2 line to R lines was reasonably induced by both the zinc deficiency and different crystal field effect upon Cr³⁺ [12,15]. Basavaraju et al. further revealed that the degree of cationic site inversion would be enhanced by the magnesium vacancy through the refinement of cationic site [11]. In addition, Jeong et al. reported that cationic site inversion in the ZnGa₂O₄ host lattice must be established when annealed in an oxygen deficient atmosphere [9]. So it was believable that the co-doping of Si⁴⁺ in this work could favor the formation of antisite defects, which would lead to the change of the relative intensity of N2 line to R lines. The influence of Si⁴⁺ co-doping onto long-lasting performances could be estimated in terms of the afterglow decay and thermoluminescence.

3.4 Afterglow decay analysis

The afterglow spectra have been recorded for each sample under excitation by 254 nm light for 5 minutes. Figure 5 displayed the four afterglow spectra corresponding to x = 0.01, 0, 0.03, and 0.05 mol of Si⁴⁺ concentration, respectively. At the first glance and by rough comparison, the persistent luminescence features keep extremely similar to their photostimulated emission spectra (see Fig. 4). The similarity in the position, shape and bandwidth indicated that the afterglow spectra originated from the same characteristic emission of Cr³⁺, in which the R lines and N2 line could be still resolved although the resolution was low. This demonstrated that two kinds of Cr³⁺ ions, at either “ideal” or distorted coordination, can both act as afterglow centers and contribute to the persistent luminescence spectra. The intensities of R lines (with respect to the intensity of N2 line) went down gradually and they became poorly resolved along with the increasing Si⁴⁺ content. But instead, the N2 line becomes stronger. The phosphor has the best afterglow emission with the resolvable R and N peaks when doped x = 0.01 mol Si⁴⁺ content.

 

Fig. 5 The persistent luminescence of samples with different co-doped Si⁴⁺ concentrations at λ_ex = 254 nm excitation.

Download Full Size | PDF

The afterglow decay curves were recorded by monitoring the overall intensity of afterglow under the ultraviolet excitation for 5 minutes. The curves were plotted in Fig. 6 sorted by the Si⁴⁺ concentrations. In general, the afterglow undergoes a complicated exponential decay containing a rapid decay process at first followed by a slow decay [23].

 

Fig. 6 The Afterglow decay curves for all ZnGa₂O₄: 0.5% Cr³⁺, xSi⁴⁺ phosphors.

Download Full Size | PDF

As shown, the persistent intensity experiences a sharp decreasing at the very beginning and then a slow decay. The second slow decay process denotes the persistence duration. A double exponential function is applied to fit the afterglow curves and the function is as follows,

Table 1. The time constants (τ_1,τ₂) fitted from the afterglow decay curves for ZnGa₂O₄: 0.5% Cr³⁺, xSi⁴⁺

View Table | View all tables in this article

3.5 Thermoluminescence

Figure 7 displayed the TL curves of phosphors. The TL curves are clearly asymmetric broad band covering the temperature range of 50 – 150 °C except that the TL of sample with 5 mol% Si⁴⁺ has one more band at higher edge. The deconvolution was carried out using the Lorentzian function to the TL curve, and the fitting gave the main TL peak at around 75°C for all phosphors (See the inset in Fig. 7). As well known, the traps play a decisive role to the long-lasting afterglow of phosphors, namely in terms of the trap depth and density corresponding to the peak position and relative intensity in TL curve. After the storage of excitation source, charge carriers captured by traps will be released to activators gradually by the thermal motion at room temperature, leading to the long afterglow emission through the recombination of electrons and holes. The electrons trapped in shallow traps can easily release to recombine with holes resulting in a short afterglow lifespan and the initial persistent intensity, and in contrast electrons in the deep traps need high activation energy to escape leading to the poor afterglow performance [24]. In hence, the abundant traps with suitable trap depth are critical and favorable to good afterglow performance. As seen from Fig. 7, all the TL curves of the phosphors have relatively large integrated area centering at 75°C and cover a temperature range of 50~100 °C, satisfying the conditions to the good afterglow performance predicted in persistent phosphors [25,26]. By the comparison of TL curves, there is only one trap at 75°C for samples with x = 0 and 0.01 Si⁴⁺-doping, while at least two traps for those with 0.05 Si⁴⁺-doping. And also the intensities of the former two samples are stronger leading to the initially large intensities of persistent luminescence. As for samples with x = 0.03, the TL presents significant difference having a peak-shifting to high temperature and a bigger spectral width. The TL peak shifted to 82°C and the intensity was lower than those of 0.01 or 0.03 samples. The TL curves were in good agreement with the initial intensities of afterglow in Fig. 6. When linking the TL curves of the samples of 0.03 and 0.05mol Si⁴⁺ together, it seemed that the TL of the sample of 0.03 Si⁴⁺ almost was in a medium shape linking from one TL peak to two peaks, corresponding to one trap and two traps in the phosphors. In the case of 0.03mol Si⁴⁺, The TL curve was broader and shifted with the trend of the formation of new trap. When the Si⁴⁺ content increased to be 0.05mol, the TL curve fully separated into two peaks and the new trap at about 130°C formed. The new trap is deeper and could explain why the sample of 0.05mol Si⁴⁺ had long persistent duration (τ₂) and better long-lasting luminescent performance.

 

Fig. 7 The TL spectra of the as-prepared ZnGa₂O₄: 0.5% Cr³⁺ phosphors with each concentration of Si⁴⁺ doping. The inset presents the fitting of the first TL peak at nearly identical T_m = 75°C.

Download Full Size | PDF

4. Effect of Si⁴⁺ co-doping

As described by Bessière and Zhuang, Zn vacancies created in the ZnGa₂O₄:Cr³⁺ phosphors take the main responsibility for TL peak [13,15]. The fitted TL peaks were placed in the inset of Fig. 7, illustrating the peak value (T_m) and the peak variation. According to approximation of the trap depth with respect to the TL peak (T_m), the depths of traps were evaluated and presented in Table 2. It is notably that when the Si⁴⁺ content changed from 0 to 0.01mol, the phosphor has the highest TL intensity, while for the phosphor with 3 mol% Si⁴⁺, the TL peak shifted to 82 °C leading to a deeper trap (0.71eV) comparing to the peaks of 75°C. The deeper trap, broader area, and fast decay rate demonstrated that the sample with 0.03mol Si⁴⁺ owned poor persistence luminescence.

Table 2. The TL peak (T_m) and the corresponding depth of trap levels for ZnGa₂O₄: 0.5% Cr³⁺, xSi⁴

View Table | View all tables in this article

Based on the study by Kang et al. and the above analysis, it could be proposed that the doping of Si⁴⁺ ions should preferentially occupy the Zn vacancies at first and then substitute for Ga³⁺ at B-sites during the crystalline proceeding [27]. In our experiments, phosphors were synthesized by the conventional high temperature method, and the doping of SiO₂ favors the formation of pure crystal phase and acts as a flux. Although the Si⁴⁺ caused the differences in both the sintered time and temperature, all phosphors still presented the persistent luminescence even for sample without Si⁴⁺. It suggested that the Zn vacancies have been indeed produced to a high degree in the phosphor without Si⁴⁺. This could be supported by the N2 line emission of Cr³⁺, which came from the Cr³⁺ affected by the antisite defects from the inversion coordination. When Si⁴⁺ was co-doped into the host, the tetravalent Si⁴⁺ ions firstly and fast took the Zn vacancies and broke the charge equilibrium, so the substitution of Ga³⁺ ions was complicated and not only the Si⁴⁺ ions but also the divalent Zn²⁺ were able to substitute for the trivalent Ga³⁺ ions. Due to the break of the charge balance and inversion coordination, the Zn vacancies increased at the beginning of the doping of Si⁴⁺, which caused the rising of the density of the trap levels from the Zn vacancies. So the persistent luminescence presented the stronger TL intensity and slower decay. However, when Si⁴⁺ content continued to increase up to 0.03 or 0.05mol, the more Zn vacancies were taken by Si⁴⁺ ions, the demand of the charge equilibrium (Zn²⁺-2Ga³⁺-Si⁴⁺) and inversion mechanism may make Zn vacancies saturated. The heavy doping of Si⁴⁺ (0.03) and the combination of two substitutions decrease the density of the first trap indicating that the persistent luminescence becomes poor. Along with the continuing increasing of Si⁴⁺ ions(up to 0.05mol), the occupation of the Ga³⁺ vacancies and inversion substitution lead to more antisite defects and the distorted octahedral sites, which were surrounded by octahedral Si⁴⁺ positive defects and induced new deeper traps (at 130°C). The deeper traps probably benefitted to the persistent luminescence performance in the case of 0.05 Si-doped. As previously mentioned, the Zn vacancies are favorable for the generation of antisite defects in the ZnGa₂O₄ host. On one hand, when the Zn vacancies decreased along with Si co-doping, and then the further occupation of Ga³⁺ by Si⁴⁺ still increased. The charge disequilibrium possibly favors the exchange of Zn²⁺ and Ga³⁺ ions on the other hand. The antisites defects enhanced the N2 line emission of Cr³⁺ and suppressed its R lines emission.

5. Conclusion

ZnGa₂O4:Cr³⁺ phosphors co-doped with x mol% Si⁴⁺ (x = 0, 1, 3, and 5) have been synthesized by using the high temperature solid-state reaction. The influence of the co-doped Si⁴⁺ concentration on their photoluminescence (PL), afterglow decay, and thermoluminescence (TL) were investigated in detail. The phosphors with co-doped Si⁴⁺ can be effectively activated by UV excitation can induce 10 minutes of reddish persistence luminescence originated from the emission of Cr³⁺. The relative intensity of N2 line and R lines of Cr³⁺ emission increases with the rising of co-doping Si⁴⁺ due to the increase of the antisite defects. The TL curves analysis indicated that the phosphor with 1 mol% Si⁴⁺ has the optimal persistent performance having positive effects to the density and depth of traps levels. Based on the change of afterglow decay and TL curves, the Zn vacancies increase at first and then decrease when the Si⁴⁺ concentrations surpass 0.01mol. The further substitutions Ga³⁺ sites by Si⁴⁺ and the charge disequilibrium favor the exchange of Zn²⁺ and Ga³⁺ causing more antisite defects which enhanced the N2 line emission of Cr³⁺ and suppressed its R lines emission.