1. Introduction

Long persistent luminescence (LPL) phosphor, as a special kind of luminescent material, is defined as the emission of materials remaining visible for an appreciable time-from seconds to hours-after the removal of the excitation source (such as UV, VUV, X-ray, γ-ray, visible light or other ionizing radiations). This phenomenon has fascinated researchers and the materials showing LPL characteristic are called long persistent phosphors (LPPs) [1–4]. Because of their unique sustainable luminescence, much attention has been focused on the exploration of novel LPPs. LPPs can find various potential applications in many important fields, e.g. emergency lighting displays, weak illumination, storage media, detection of high-energy rays such as X-rays and β-rays, fiber-optic thermometers and in vivo bio-imaging [5–9], etc.

At present, it is generally believed that the LPL generation is attributed to the following processes: electrons or holes can be trapped and de-trapped by or from the valence band or conduction band under excitation. Then, the trapped electrons and holes in these traps also can be thermally released at room temperature after the removal of excitation, leading to LPL phenomenon [10–12]. So, two prerequisites seem to be necessary for the better LPL performance: one is the selection of a suitable host and another is luminescent center. Based on this principle, researchers in this field have synthesized various kinds of LPPs with different activators in different lattice hosts. In order to search for new LPPs with better LPL properties and further understand the LPL mechanisms, numerous of papers have been published. However, the traps nature and their mechanisms of capturing energy are complicated and not understood completely. The last several decades have witnessed a faster and faster development of rare-earth ion doped LPPs [13–15]. In recent years, Cr³⁺-doped LPPs with near infrared LPL have been continuously reported [8, 16]. In contrast, few studies have been reported on the transition metal ion (Mn²⁺) doped germanates based LPPs. We found that zinc or magnesium containing germanates are suitable for Mn²⁺ ion doping. Mn²⁺-doped zinc or magnesium containing germanates have attracted interests due to their potential applications in field-emission displays. Mn²⁺ ion is typical of a d⁵ electron configuration. The emission is strongly influenced by the crystal field. It shows green emission in tetrahedral symmetry and red emission in octahedral symmetry [17]. At present, some studies on Mn²⁺-doped germanates with LPL have been reported, including Na₂ZnGeO₄:Mn²⁺, Li₂ZnGeO₄:Mn²⁺, CaZnGe₂O₆:Mn²⁺, Zn₂GeO₄:Mn²⁺ and Mn²⁺-doped MgO-GeO₂ [10, 18–21]. It infers that zinc or magnesium containing germanates may be also suitable hosts for Mn²⁺ ion to fabricate new LPPs.

Therefore, in this paper, we introduced the preparation of Li₂MgGeO₄ and Li₂MgGeO₄:Mn²⁺ phosphors: non-doped and Mn²⁺-doped Li₂MgGeO₄ phosphors were synthesized successfully by a high temperature solid-state reaction method. As previously expected, green LPL from Li₂MgGeO₄:Mn²⁺ was observed after the excitation by 254 nm. The photoluminescent properties of Li₂MgGeO₄ and Li₂MgGeO₄:Mn²⁺ were studied in detail based on steady spectroscopy. With the aid of LPL decay and thermoluminescence (TL) properties, LPL mechanism was discussed.

2. Experimental

2.1 Synthesis

A series of non- and Mn²⁺-doped Li₂MgGeO₄ (LMG) powder phosphors were synthesized via a high temperature solid-state reaction method. Li₂CO₃ (AR), MgO (AR), MnCO₃ (AR) and GeO₂ (99.99%) were used as raw materials. Stoichiometric amounts of starting reagents according to the nominal compositions of Li₂Mg_1-xGeO₄: xMn²⁺ (x = 0 and 0.001, denoted as LMG and LMGM, respectively) were milled thoroughly in an agate mortar. The mixtures were fired at 1200 °C for 5 h in a tube furnace. Finally, the obtained samples were ground again to fine powders for the subsequent characterization.

2.2 Measurement

The X-ray powder diffraction patterns were obtained by a XD-2 powder diffractometer (Beijing PGENERAL) using Cu Kα irradiation (λ = 1.5406 Å) under 36 kV tube voltage and 20 mA tube current with a scanning step of 0.02° in the 2θ range from 10° to 70°. The reflectance was measured by a UV–visible spectrophotometer (Shimadzu UV-2450, Japan) using BaSO₄ as a reference. The fluorescence spectra and LPL spectrum were measured using a FLS-980 Fluorescence Spectrophotometer (Edinburgh instruments). The LPL decay curves were recorded by a GFZF-2A single-photon counter system after the samples were irradiated by 254 nm for 1 min. TL curves were collected by a FJ427A1 thermoluminescent dosimeter (CNNC Beijing Nuclear Instrument Factory). The samples were pre-irradiated by a UV lamp (~254 nm) for 1 min with a waiting time of 2 min and then heated at a linear rate of 1 °C/s.

3. Results and discussion

3.1 Phase structure

The JCPDS card No. 24-0627 and the X-ray diffraction patterns of the as-obtained samples (LMG and LMGM) are shown in Fig. 1. It is obvious that all diffraction peaks match with the standard data of LMG (JCPDS No. 24-0627) except a tiny of unknown impurity (marked by blue cycles). It can be concluded that the as-obtained samples are structurally and chemically LMG. The incorporation of a small amount of Mn²⁺ ions into LMG host lattice almost has no significant change on the general structure. LMG crystallized into an orthorhombic crystal system with the space group Pmn21 (31) and the unit cell parameters: a = 6.40, b = 5.48, c = 5.00 Å. Mg²⁺ is surrounded by four oxygen ions forming a regular tetrahedron. Owning to the similar radii and valence between Mn²⁺ (0.66 Å, CN = 4) and Mg²⁺ (0.57 Å, CN = 4), Mn²⁺ ions tend to occupy Mg²⁺ sites.

 

Fig. 1 XRD patterns of sample LMG:xMn (x = 0 and 0.001) and that of standard data of Li₂MgGeO₄ (JCPDS 24-0627) given for comparison.

Download Full Size | PDF

3.2 UV-visible diffuse reflectance spectra

Figure 2 (a) exhibits the diffuse reflection spectra of the undoped and Mn²⁺-doped LMG samples. The absorption in the range below 240 nm is attributed to the UV absorption of LMG host. It is notable that the diffuse reflectance spectrum of LMGM shows a near plateau of high reflection in the wavelength range of 800-550 nm and begins to decrease from 550 nm. The absorption in the region of 550–350 nm is caused by d–d transitions of Mn²⁺ and the absorption in the region of 350–240 nm is assigned to the charge transfer state (CTS) of O²⁻–Mn²⁺ transitions .

 

Fig. 2 (a) Diffuse reflectance spectrum of sample LMG and LMGM ; (b) The function of [F(R)hλ] versus photon energy hλ for sample LMG.

Download Full Size | PDF

According to Kubelka-Munk function (1), the optical band gap of LMG host can be estimated [22].

3.3 Photoluminescence and persistent luminescence properties

Figure 3(a) shows the excitation and emission spectra of undoped LMG. When monitoring emission at 350 nm, a strong broad excitation band in the wavelength range below 250 nm and a weaker excitation band in the wavelength range between 250 and 300 nm can be observed. Undoped LMG shows a nonstructure broad band emission between 300 and 550 nm under excitation by 220 nm. It can be ascribed to the recombination of donors ($V_O^{••}$ and $M{g}_i{}^{••}$) and acceptors ($V_{Ge}^{\prime \prime \prime {\prime }^{}}$ and $V_{Mg}^{\prime \prime }$) which could be created as intrinsic defects in host lattice in the process of high temperature solid-state reaction.

 

Fig. 3 (a) The excitation and emission spectra of sample LMG; (b) The excitation and emission spectra of sample LMGM.

Download Full Size | PDF

The excitation and emission spectra of sample LMGM are shown in Fig. 3(b). Monitoring emission at 532 nm, a broad excitation band in the range from 210 to 350 nm peaking at 250 nm which is mainly attributed to the transition of electrons in 2p⁶ orbital of O²⁻ to 3d⁵ orbital of Mn²⁺ is observed. It is overlapped with the LMG host excitation spectrum. Besides, several sharp and weak peaks in the wavelength range of 350-500 nm originate from the 3d-3d transition of Mn²⁺ ions [24]. When excited by 250 nm, the emission spectrum consists of two broad emission bands: one is in the wavelength range of 300-500 nm which is assigned to the LMG host emission; another is a strong emission band between 500 and 600 nm with a maximum at 532 nm which can be attributed to the ⁴T₁(⁴G)-⁶A₁(⁶S) transition in Mn²⁺ ions. The green emission of Mn²⁺ confirms that the coordination environment of Mn²⁺ in the LMG host lattice is tetrahedral symmetry (i.e. Mg sites). Evidently, the LMG host emission overlaps with the excitation of Mn²⁺ caused by 3d-3d transitions, which infers the energy transfer from host to Mn²⁺.

Figure 4(a) presents the LPL spectrum of sample LMGM collected immediately after the 254 nm excitation source was switched off. The bright green LPL which mainly originates from Mn²⁺ emission can be observed by human naked eyes in dark after after irradiation with short wavelength (~254 nm) rather than long wavelength. As shown in Fig. 4(b), the LPL intensity decreases quickly at first and then decays slowly. After being excited by 254 nm for 10 min, the green LPL of LMGM can last more than 5 h.

 

Fig. 4 (a) LPL spectrum of sample LMGM after the removal of UV light source (254 nm); (b) LPL decay curve of sample LMGM.

Download Full Size | PDF

3.4 Thermoluminescence properties

Though numerous of studies on LPPs have been carried out, a general mechanism that suits all LPPs still has not been known because that the nature of the traps and the processes of capture and releasing energy are too complicated to be understood completely at present [3, 25]. It is commonly believed that the stored energy transfer from traps to traps or to emission centers causes the LPL. LPL property strongly depends on the traps in the host, i.e., materials have traps with suitable depth and high density are likely to show good LPL performance.

Since that the LPL phenomenon results from energy transfer from traps in LPPs, a TL technique was used to obtain useful information about the trap distribution and the interactions between them. In order to characterize the traps in the as-obtained samples, TL measurements were performed on LMG and LMGM. Before the measurements, the samples were irradiated using a 254 nm UV lamp for 1 min and then put in dark for 2 min. Figure 5(a) show the TL curves of LMG and LMGM phosphors. Exactly, the TL intensity of LMG is very weak, which demonstrates that no LPL can be observed from undoped LMG phosphor. The weak TL curve may be caused by the intrinsic defects such as $V_{Mg}^{\prime \prime }$, $V_{Ge}^{\prime \prime \prime {\prime }^{}}$, $V_O^{••}$, $M{g}_i{}^{••}$. LMGM show a strong and broad asymmetry TL curve with a maximum ~110 °C. Compared with the TL curve of LMG, one can consider that the green LPL in LMGM is also controlled by the intrinsic defects. It is no doubt that the equivalent substitution of Mg²⁺ constituent by Mn²⁺ creates isoelectronic traps and the creation of intrinsic defects ($V_O^{••}$, $M{g}_i{}^{••}$) is largely promoted. The incorporation of Mn²⁺ into LMG host lattice significantly enriches the traps. The positive defects can serve as electron traps while negative defects can act as hole traps. It is too broad (from 30 to 300 °C) to be considered as a single TL peak as it looks. Moreover, the TL curves recorded at different delay times are shown in Fig. 5(b). It shows a significant shift to high temperature with longer delay time. Therefore, we assume that the existence of several TL peaks distribution between 30 and 300 °C form the broad TL curve. To resolve the TL curve and determine their kinetic parameters of individual peaks, the deconvolution of the TL curve was carried out.

 

Fig. 5 (a) TL glow curves of samples LMG and LMGM; (b) TL glow curves of sample LMGM recorded at different delay times; (c) The fitting results for sample LMGM.

Download Full Size | PDF

According to Eq. (3) provided by McKeever [26]. The TL curve was well deconvoluted into four individual TL peaks with the maximum at 70, 98, 135 and 186 °C (denoted as peak 1 to peak 4), as shown in Fig. 5(b).

The key parameters including trap depth (E) and kinetics order (b) as well as the temperature of peak maximum corresponding to individual peak are listed in Table 1. Note that the values of b ~2 for four TL peaks suggest a second kinetics order. So the released energy from shallow traps has a great possibility to be recaptured by deep traps, which is benefit for longer LPL duration. LPL in LMGM phosphor is related to traps (TL peaks 1-3) with depths (0.69-0.82 eV) rather than TL peak 4 because that the depth of TL peak 4 is too deep to have any contribution to LPL.

Table 1. Fitted results including T_m, E and b of TL glow curve for sample LMGM.

View Table

3.5 Possible LPL mechanism

Based on the aforementioned results, LPL mechanism of Mn²⁺ doped LMG can be qualitatively summarized as follows. Under the UV (254 nm) excitation, electrons are excited from the ground state of Mn²⁺ to the conduction band. The recombination of the excited electrons and holes generates the 350 nm self-activated luminescence. There is also a possibility of energy transfer from host to Mn²⁺ ions. In addition, after non-radiative relaxation to ⁴T₁(⁴G) level, the majority of electrons will jump to ground state ⁶A₁(⁶S) and generate the green emission of Mn²⁺. The residual excited electrons may be captured by electron traps just below the conduction band. Meanwhile, the holes near the valence band can be captured by hole traps. After the removal of irradiation source, the trapped electrons and holes can be thermally released at a slow rate at room temperature. Then, the recombination of the released electrons and holes leads to the green LPL.

4. Conclusions

In conclusion, a novel phosphor Li₂MgGeO₄:Mn²⁺ exhibiting green emission has been synthesized via a high temperature solid-state reaction method. The photoluminescence and LPL properties were characterized. It is revealed that the green photoluminescence and LPL originate from Mn²⁺ ions. The green LPL can maintain for a time exceeding 5 h after the removal of 254 nm excitation source. Furthermore, the traps generated in host lattice and the possible LPL mechanism were discussed in detail.