1. Introduction

Long afterglow refers to luminescence that can persist for hours after the excitation is stopped [1]. It has been confirmed to have rich applications in emergency, luminous paints, pressure sensor, vivo imaging, medical applications, etc [2]. Generally, afterglows are created by radiative transitions of trapped electrons releasing from trap centers, where the electrons could continuously release and move to luminescence centers. The process could last a long period resulting in continuous luminescence [3]. In the past decades, numerous inorganic materials including aluminates, phosphates, titanates, silicates, etc have been developed as afterglow phosphors [4]. To dope rare earth ions such as Dy³⁺, Tb³⁺, Eu^3+/2+, Pr³⁺ etc in a phosphor has been confirmed to be an active method to improve afterglow performances [5].

In the present work, afterglow on BaZr₄(PO₄)₆ was developed on several motivations. Firstly, as one of the closed-shell transition-metal complexes, Zr⁴⁺ ion acts as an efficient luminescence center when it locates in an asymmetric site. Fig. 1(a) is the schematic picture of ZrO₆ in BaZr₄(PO₄)₆. Zr-O in Zr(1)O₆ and Zr(2)O₆ has changes in the regions of 1.9375-2.2578 Å and 1.9687-2.1852 Å, respectively. The distortion induces the luminescence from Zr⁴⁺. Secondly, except for the luminescence centers, afterglow was reported in Zr⁴⁺-containing hosts. Zr⁴⁺ ion could play the role of electron traps in the lattices. Importantly, Zr⁴⁺-associated traps have a suitable depth for the improvement of afterglow phosphors [6]. Thirdly, it is well-known that the rich defects such as cation or anion vacancies are very important for the improvement of afterglow. BaZr₄(PO₄)₆ (Ba_0.5Zr₂[PO₄]₃) has a triclinic structure on the base of NaZr₂[PO₄]₃ by changing the Na⁺ by 0.5Ba²⁺ [7]. As noted in Fig. 1(b), there are great deals of cation vacant (V_Ba) in the lattices, which could give possible intrinsic defects. Moreover, the structure presents excellent flexibility, which is favorite for the numerous ionic substitutions in the lattice sites. This is favorable for the creating of the extrinsic defects by aliovalent ion doping or preparing the samples in different atmospheres.

 

Fig. 1 (a): surrounding of Zr(1)O₆ and Zr(2)O₆; (b): schematic view along [100]; (c): the structural refinement; (d): SEM picture of BaZr₄(PO₄)₆.

Download Full Size | PDF

The BaZr₄(PO₄)₆ phosphor was synthesized by solid-state reaction method. XRD refinement, PL spectra were measured. The afterglow was detected after removal of UV irradiation, which could be enhanced by Eu³⁺ doping and sintering in a CO-atmosphere. The trap mature and its depth were investigated by TL spectra. The mechanism was also discussed.

2. Experimental

Ba_1-xEu_xZr₄(PO₄)₆ (x = 0, 0.1) was prepared via solid-state reaction. The raw chemicals are stoichiometric BaCO₃ (A.R.), (NH₄)₂HPO₄ (A.R.), ZrO₂ (A.R.), Eu₂O₃ (99.9%). The reactions were taken placed in three steps. At first, the mixture was pre-heated at 300 °C for 3 h in air. Secondly, the powder was thoroughly grinded and then sintered at 950 °C for 4 h in air. At last, the mixtures were mixed and sintered at 1300 °C for 10 h in air and CO atmosphere. The CO-sintering was realized by burying the sample-contained crucible in active carbon which was fully filled in a big crucible with a close lid. That is to say, four samples were prepared, i.e., the pure- and Eu³⁺-doped BaZr₄(PO₄)₆ sintered in air- and CO-atmospheres.

The XRD measurements were finished on the diffractometer (Rigaku D/Max, 40 kV, 30 mA) equipped by Bragg-Brentano geometry using Cu-Kα radiation (λ = 1.5405 Å). The PL excitation and emission spectra were conducted by the luminescence spectrometer (Perkin-Elmer LS-50B), which is equipped by Monk–Gillieson type monochromators and a xenon lamp as excitation. Thermoluminiscence (TL) signals were recorded on the TL meter (FJ-427A, Beijing Nuclear Instrument Factory China). Before the TL measurements, the sample was exposed to a UV light for 10 minutes. The heating rate of 2 °C/second was selected.

3. Results and discussions

3.1 The phase formation and structure

The phase formation of BaZr₄(PO₄)₆ powders was confirmed by XRD measurements. The patterns were completed by structural refinement in the GSAS program (Fig. 1(c)). The refined data indicate that the structure has a trigonal space group of R-3 (148) with the unit cell of a = 8.6975(11) Å, c = 24.007(15) Å, V = 1572.56(32) ų, and Z = 3. The good refinement (R_wp = 0.1157) indicates that the structure has a high crystalline purity.

Figure 1(b) shows the structure sketch a unit cell along [100] modeled by the atomic coordinate’s in refinement. The three-dimensional framework is constructed by PO₄ tetrahedra and ZrO₆ octahedral by corner-sharing. This network develops a rigid and highly stable lattice with four structural tunnels per formula unit; However the tunnels are only partially occupied by Ba ions, resulting in a large numbers of vacant Ba, V_Ba. The microstructure of phosphor surfaces are shown by SEM photo in Fig. 1(d). In this solid-state preparation at 1300 °C the particle size varied from 0.3 to 1.5μm. The surfaces of the particles are smooth and clean indicating a good crystallization.

3.2 The photoluminescence spectra

The PL spectra of pure and Eu-doped BaZr₄(PO₄)₆ prepared in different atmospheres are displayed in Figs. 2(a)-2(d). BaZr₄(PO₄)₆ prepared in air displays broad emission peaked at 475 nm under 313 nm excitation (Fig. 2(a)). The CIE color coordinates were calculated to be (x = 0.231, y = 0.312), which is close to the white region. This emission could be assigned to the well-known charge-transfer (CT) transition, i.e. an electron excited from oxygen into empty 4d orbital of Zr⁴⁺. However, Zr⁴⁺ luminescence by CT transition has been widely reported in UV region such as YPO₄:Zr (290 nm) [8], ZrP₂O₇ (285 nm) [9], Sr₂ZrO₄ (320 nm) [10], BaHf_1-xZr_x(PO₄)₂ (356 nm) [11], BaZrSi₃O₉ (285 nm) [12], or even in VUV region such as BaZr(BO₃)₂:Eu (180-200 nm) [13]. Furthermore, as reported in ZrO₂ [14], this emission also has a possibility from Ti³⁺ impurities since almost all Zr sources have some Ti impurities_. In this work, this possibility could be excluded because this band was quenched in both pure and Eu³⁺-activated samples prepared in CO atmosphere (Figs. 2(b),2(d)). Actually unusual luminescence of Zr⁴⁺ in visible region is possible, for example, Ca₄ZrGe₃O₁₂ (540nm) [15], SrMg₂P₂O₈:Zr (500nm) [16], β-Zn₃(PO₄)₂:Zr⁴⁺ (485nm) [17]. The low energy emission has been suggested due to the unusual disordered coordination environment of Zr⁴⁺ ions [17]. In BaZr₄(PO₄)₆, ZrO₆ environment is greatly disordered. This could be reflected by the fact that the Zr-O distance in Zr(1)O₆ and Zr(2)O₆ has great changes in the regions of 1.9375-2.2578 Å and 1.9687-0.1852 Å, respectively. The distortion induces the luminescence from Zr⁴⁺ ions.

 

Fig. 2 The PL spectra of pure BaZr₄(PO₄)₆ (a, b) and the Eu-doped samples (c, d) prepared in different atmospheres.

Download Full Size | PDF

On the emission spectra (λ_ex = 313nm) for Ba_0.9Eu_0.1Zr₄(PO₄)₆ prepared in air, the lines from Eu³⁺ can be detected together with Zr⁴⁺ emission (Fig. 2(c)). The 470nm excitation has a maximum at 313 nm. The excitation of the ⁵D₀→⁷F₂ (615nm of Eu³⁺) consist of a weak band at 275nm from CT transition in Eu³⁺-O²⁻ and the 4f–4f transitions in Eu³⁺. As noted in Fig. 2(c) there is limited overlap between the excitation of Eu³⁺ and the host emission. The spectral information indicates that the energy transfer from the host to Eu³⁺ in BaZr₄(PO₄)₆ is weak. As shown in Fig. 2(d), the excitation band and emission of Ba_0.9Eu_0.1Zr₄(PO₄)₆ prepared in CO have great changes. The peak emission from Eu³⁺ disappeared, indicating that Eu²⁺ ions were reduced. The emission could be assigned to the 5d to 4f transitions in the lattices. The luminescence differences in two samples reflect in the different coordination surrounding of Zr⁴⁺.

3.3 The long afterglow

The afterglow characteristics are shown in Fig. 3. The pale-white can be seen with the naked eye in the dark after the excitation. The afterglow decay present at least into two parts, one is at the initial part with rapid decay, and another is in the detail with slow process. This indicates the existence of complicated hole- or electron-traps. Both Eu-doping and CO-sintering could enhance the long-lasting time of BaZr₄(PO₄)₆. The afterglow intensity of the CO-sintered phosphor is rather stronger than that of the air-sintered samples. It is reasonable that the oxygen vacancies can influence the afterglow characteristics by acting as the electron trap. This indicates that both the doping and CO-heating create more defects in lattices, which acts as trapping centers and has a significant influence on afterglow. To characterize the traps in our samples, TL measurements were performed and illustrated in Fig. 4.

 

Fig. 3 The afterglow decay curves of pure and Eu³⁺-doped BaZr₄(PO₄)₆ prepared in different atmospheres.

Download Full Size | PDF

 

Fig. 4 The thermoluminsecent spectra of pure and Eu³⁺-doped BaZr₄(PO₄)₆ prepared in different atmospheres as noted in the figure.

Download Full Size | PDF

3.4 TL signals

For BaZr₄(PO₄)₆ prepared in air, only one peak located at 94 °C (367 K) is observed (Fig. 4(a)). This peak was observed in the four samples except the shift to low temperature. The trap could be assigned to the intrinsic defects associated to Zr⁴⁺ centers labeled as T_A(Zr). Similar Zr⁴⁺-related traps have been reported in afterglow phosphor such as β-Zn₃(PO₄)₂:Mn²⁺,Zr⁴⁺ (344 K) [18] and Zr⁴⁺-doped SrMg₂(PO₄)₂ (368 K) [16]. The extra TL peak appears at 153 °C (426 K) together with the increase of the TL intensity (Fig. 4(b)) can be observed after the Eu³⁺ doping. This indicates that Eu³⁺ doping improves the number of trapping centers. The trivalent Eu³⁺ ions will substitute Ba²⁺ creating the positive defect of [Eu_Ba]^·. V_Ba” vacancy should be created to keep the charge balance. The trap is notes as T_B (Eu_Ba^·) corresponding to the peak at 153°C (426 K). This indicates that the Eu³⁺ ions act as a trap center rather than an emission one. Compared to the peaks at 94 °C, the one at 153 °C is more difficult to liberate trapped carriers due to the higher activated energy. Thus electron traps may have a small contribution to practical relevance in afterglow at room temperature.

It is found that the afterglow performance is improved when the samples were sintered in CO-atmosphere. On the TL of pure (Fig. 4(c)) and Eu-doped BaZr₄(PO₄)₆ (Fig. 4(d)), another TL peak at 110 °C (383 K) can be observed. This could be ascribed to oxygen vacancies (VO) that usually appear in the oxide host when a sample is sintered at reducing atmosphere [19]. This trap could be labeled as T_C (V_O^··). It could be observed that T_A (Zr) has a shift to low temperature from TL curve (a) to (b), (c), and (d). The trap depth can be expressed by the Eq. (1) and (2) [20].

When a carrier is trapped at temperature T, the dynamic process can be expressed by$p={\tau }^{-1}=s\times \exp (-E/k{T}_A)$, τ is decay time, s is constant frequency factor. E is trap depth, and p is probability for a carrier escapes from trap [21]. It is obvious that if traps are too deep for a trapped electron, it will give a small probability for the releasing. Suitable TL peak should locate in 320-380 K for an easy release of an electron [5]. It can be found that the TL temperatures at T_A and T_c traps have a low depth for the possible application. As shown in Fig. 4, the sintering in CO reduction shifts the TL spectra of both pure and Eu³⁺-activated samples to low temperature. This could be due to the second order kinetic since in this process, traps with the same energy but lower concentration are shifted to higher temperature [[22]].

3.4 The afterglow mechanism

On the basis of the above-mentioned results, a possible mechanism is proposed to explain the afterglow generation in BaZr₄(PO₄)₆ as in Fig. 5. The excitons can be induced by UV- or near-UV irradiation from CT transition in Zr⁴⁺–O²⁻ (step 1). After that the electron–hole pairs will be separated. The holes could be trapped in V_Ba (step 2), while some electrons could move into the tunnel state (step 3) and parts of them go to the conduction band. Then the induced electrons could be trapped at T_A, T_B, and T_C centers (step 4). When the temperature excesses the trap depth, the electrons will thermally release into the tunnel state (step 5) following by transferring to excited level of Zr⁴⁺ ions (step 6), and relax and finally emit the pale white (Zr⁴⁺) afterglow (step 7). In this process, Eu³⁺ doping does not produce afterglow on itself. The role of Eu³⁺ doping on Ba²⁺ only induces the rich traps by creating the positive defect of [Eu_Ba]^·. The Eu-doping and sintering in CO-atmospheres could modify the luminescence color and prolong the afterglow time.

 

Fig. 5 The mechanism of afterglow decay of pure and Eu³⁺-doped BaZr₄(PO₄)₆ prepared in different atmospheres.

Download Full Size | PDF

As shown above, the determination of TL curves displayed that CO-sintering will significantly increase the oxygen vacancies to hold electrons. More oxygen vacancies mean that more electrons on excited levels will be trapped. As the result, more excited electrons will be delayed to jump back to the ground states to extend the afterglow time of phosphors.

4. Conclusions

In conclusion, long afterglow phosphor BaZr₄(PO₄)₆ was synthesized by solid state reaction method. The structural refinement indicates that the phosphate bears plenty of cation vacancy and space in the lattices. Both pure- and Eu-doped BaZr₄(PO₄)₆ present efficient self-activated luminescence due to CT transitions around 475 nm. Both Eu-doping and CO-sintering could enhance the long-lasting time. Comparing with the air-sintered sample, a significant enhancement in the afterglow for the CO-sintered sample was observed. This improvement could be attributed to the increase of oxygen vacancies which act as the sensitizer and the electron traps. Sintering in reducing atmosphere is a very active method to modify the afterglow in this Zr-containing phosphate. The defects act as traps were investigated with thermoluminescence. This is a promising host on afterglow phosphor by considering kinds of doping in the lattices.