Design, synthesis and characterization of a novel bluish-green long-lasting phosphorescence phosphor BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺

Abstract

Through drawing upon a solid-state reaction, a newly proposed long-lasting phosphor BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ is presented and prepared in this work. The thermoluminescence properties of the phosphor are substantially extended, and the long-lasting phosphorescence behavior is markedly intensified by virtue of the consolidation of Nd³⁺ ions which serve as trap centers. In line with density functional theory calculations, the conduction band is mostly composed of Lu 5d states while the Ba 5d states only have a tiny contribution. We analyzed the relationship between the phosphor’s electronic structure and its optical properties. The photoluminescence emission spectrum shows a blue emitting band with a wide asymmetric property and an extremum of 426 nm, arising from the 5d–4f transitions of the Eu²⁺ ions which occupy the Ba and Lu sites. It is asserted that the long-lasting phosphorescence of the Eu²⁺ ions which take up both Ba and Lu sites stems from the special form of conduction band and the occupying environment of the emission center. Yet, they have different contributions and induce an interesting phenomenon in which the blue emitting phosphor shows a bluish-green phosphorescence. The long-lasting phosphorescence can last in excess of 6 h at the recognized intensity level (0.32 mcd m⁻²) after excitation for 10 min. This work provides a new way of thinking for the development of multicolored LLP materials. This work analyzes and sheds light on the specific courses and provides a likely mechanism for the process.

---

1. Introduction

Long-lasting phosphorescence (LLP) is an optical phenomenon of great significance. Luminescence can be observed after ceasing the excitation after a couple of minutes and probably persists over a longer period of time of several hours.^1,2 Under excitation, energy can be stored and absorbed by LLP materials (inclusive of artificial light and sunlight). Subsequently, the stored and absorbed energy can be released as persistent visible light under thermal stimulation, commonly at room temperature.^3,4 LLP materials have aroused considerable attention and have been applied in extensive important fields by virtue of its prominent characteristics which include being recyclable, energy saving, and eco-friendly.⁵ Initially, LLP materials were adopted for night-time displays, traffic signs, in or on buildings, in public places, for emergency signs, and security signs.⁶ Additionally, the foregoing materials have been adopted in structural damage sensors, radiation detectors, and optical storage media.⁷ Thus far, LLP materials have been adopted in a growing number of fields to address relevant challenges, such as in in vivo bio-imaging, medical diagnostics and solar energy utilization, etc.⁸

Moreover, large numbers of LLP materials have been developed and reported. LLP materials colored red, green and blue are primarily highlighted and discussed with a predominant amount of the research reported adopting these materials for a particular purpose. Theoretically, any color material can be attained through blending these three predominant colors and changing their proportions.^9–12 Yet, implementing the mentioned approach is difficult. Although two of the tricolor LLP materials, blue (CaAl₂O₄:Eu²⁺, Nd³⁺, >24 h) and green (SrAl₂O₄:Eu²⁺, Dy³⁺, >24 h), are very bright and can be adopted for use in the market, red LLP materials are overall inadequate when adopted for a particular purpose.^13,14 Making the various components consistent with each other during the phosphorescent decay can be hard and ensuring that the color of the phosphorescence is uniform can be difficult. Furthermore, a single source of excitation can be inefficient in exciting the majority of LLP materials.^15,16

In this regard, newly proposed LLP materials consisting of multiple colors will be critically and urgently explored and developed. Currently, there is a growing interest in silicate-based luminescent materials due to their excellent chemical stability, thermal stability, weather resistance, and low temperature synthesis.^17,18 In 2010, M. Wierzbicka-Wieczorek et al. initially developed and reported the trisilicates BaRe₂Si₃O₁₀ (Re = Sc, Yb, Er, and Gd).¹⁹ The topological structure of the BaRe₂Si₃O₁₀ trisilicates consists of trisilicate Si₃O₁₀ groups in the shape of a horseshoe and ReO₆ octahedra distorted by zigzag chains of edge-sharing. Subsequently, the foregoing trisilicate compounds, e.g. BaSc₂Si₃O₁₀ and BaY₂Si₃O₁₀ are adopted to synthesize numerous versatile lanthanon doped phosphors. Additionally, the application of these phosphors in white LEDs has been investigated, as well as their photoluminescent properties.^20–24 In 2014, the bivalent Eu²⁺ doped BaM₂Si₃O₁₀:Eu²⁺ (M = Sc, Lu) phosphor was prepared and proposed by Jakoah Brgoch et al.²⁵ In 2016, Kai Li et al. reported the preparation and photoluminescent properties of the tervalent Ce³⁺ and Tb³⁺ co-doped BaLu₂Si₃O₁₀ phosphor.²⁶ The BaLu₂Si₃O₁₀ phosphor has prominent luminescence properties, which has aroused massive research interest. Until now, as indicated by a literature review, no papers have studied the LLP properties of this phosphor. Traps are significant for photo-energy storage of persistent and thermo-stimulated phosphors. Suitable trap depths and trap densities are beneficial for improving LLP properties. Co-doping Re³⁺ ions counts as one of the most frequently adopted approaches to generate new traps or update the internal properties of traps arising from nonequivalent substitution. As Dorenbos’ theory indicates that Nd³⁺ ions count as adequate traps, Nd³⁺ ions are deemed trap centers and can be incorporated into BaLu₂Si₃O₁₀:Eu²⁺ hosts.²⁷ Based on the above reasons, we have synthesized a LLP BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺phosphor. In this work, the electronic structure, photoluminescence, and phosphorescence properties of this phosphor are investigated systematically and shed light on.

2. Experimental

2.1 Synthesis

Following a frequently adopted method using high-temperatures in the solid state, this work fabricated the desired samples Ba_1−2xLu₂Si₃O₁₀:xEu²⁺, xNd³⁺ (x = 0, 0.004, 0.006, 0.008, 0.01 0.02 and 0.03, respectively presented as S1 to S6) whereby the raw materials were Eu₂O₃ (99.99%), Nd₂O₃ (99.99%), SiO₂ (99.99%), and BaCO₃ (99.99%). The relative raw materials were mixed in stoichiometric amounts and intensively ground uniformly in an agate mortar for 40 min, while incorporating an adequate amount of ethanol. The mixture was placed in alumina crucibles and sintered for 5 h at 1400 °C (N₂ : H₂ = 95 : 5) in a reducing atmosphere in a tube furnace powered by electricity. After calcining, the temperature of the samples was reduced within the furnace to room temperature and ground a second time for further application.

2.2 Characterization

These samples were characterized by powder X-ray diffraction (XRD) using a Rigaku diffractometer featuring Ni-filtered Cu Kα radiation at scanning steps of 0.02° with the 2θ value ranging from 10° to 80°. This enabled the overall phase structures of the samples to be examined. Through employing a F30 S-TWIN electron microscope (Tecnai™G2, FEI Company) high-resolution transmission electron microscopy (HRTEM) images, transmission electron microscopy (TEM) images, and an energy dispersive X-ray spectroscopy (EDX) spectrum were attained. Using a UV-Vis spectrophotometer (PE Lambda 950) and a sample of BaSO₄an ultraviolet-visible (UV-vis) diffuse reflectance spectra (DRS) was attained and used as a reference. A Xe-900 xenon arc lamp (450 W), with a FLS-920T fluorescence spectrophotometer as the source of excitation, was used to measure the luminescence decay curves, the LLP spectra, the photoluminescence excitation (PLE), and the photoluminescence emission (PL). 1 nm was established as the scanning step. 10 min after irradiating the samples with ultraviolet light a PR305 long afterglow instrument was used to measure the decay curve of the LLP. TL curves were provided using a FJ-427A TL meter (Beijing Nuclear Instrument Factory) with a heating rate of 1 K s⁻¹ and a temperature range of 20 to 400 °C. In this work, UV light was used to irradiate 0.0020 g samples which had been pressed into pellets for 5 min prior to measuring. The photographs of BaLu₂Si₃O₁₀:Eu²⁺ and BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ under/after UV irradiation were taken in a dark room by a common digital camera with certain parameter settings (exposure time, ISO value, and aperture value). With the exception of the TL curves, these measurements were conducted at room temperature. Through using the CASTEP code and abiding by density functional theory, this work calculated the electronic structure adopted by BaLu₂Si₃O₁₀. Among the theories of density function, this work selected the local-density approximations (LDA).

3. Results and discussion

3.1 XRD patterns and Rietveld refinement

XRD was used to ascertain the phase purity adopted by the prepared samples. Given that the structure of BaLu₂Si₃O₁₀ has never previously been proposed in literature, this work references the crystallographic data and initial model structure with the purpose of refining them for BaYb₂Si₃O₁₀ (ICSD 17781). The representative XRD patterns for the relative samples are presented in Fig. 1a (S1–S6). All of the diffraction peaks are used as a standard reference for the BaYb₂Si₃O₁₀ (ICSD 17781) compound.²⁶ No impure phases were detected in the attained samples, which indicates that a single phase is taken up by all of the attained samples, and the Eu²⁺ and Nd³⁺ ions are successfully incorporated into the BaLu₂Si₃O₁₀ host. The XRD pattern of the BaLu₂Si₃O₁₀ host is presented in Fig. 1b on the basis of the Rietveld structural refinement which was attained using the Materials Studio program. The background, the experimental pattern and the calculated pattern are denoted by a yellow solid line, black crosses, and a red solid line, respectively. The Bragg reflection positions taken up by the acquired pattern are indicated by short vertical pink lines. At the bottom, the blue line shows how the calculated and experimental results are different from each other. Rietveld refinement effectively examines how atoms are positioned in a primitive cell. The residual factors are R_wp = 13.78%, R_p = 9.75% and χ² = 3.127, which demonstrate that the refined results are reliable. The refinement results confirm that the compound is in the monoclinic space group P2₁/m (no. 11) with cell parameters a = 5.3541(1) Å, b = 12.1224(3) Å, and c = 6.7829(2) Å. The final refinement results indicate that this phosphor has two diverse cations, namely 8-coordinated Ba²⁺ (r = 1.42 Å) and 6-coordinated Lu³⁺ (r = 0.861 Å), presented in Fig. 1b. Distorted SiO₄ tetrahedra which form corner-sharing zig-zag trimers are contained in the crystal structure. Two Si⁴⁺ cations are given which are crystallographically independent. Through adopting two highly distorted edge-sharing LuO₆ octahedra, this work connects the foregoing SiO₄ trimers.²⁷ Generally, the radii difference between the two cations primarily determines whether the doped ions can substitute for the host cation.²⁵ The percentage difference in ionic radii between the substituted and the doped ions should be less than 30%.²⁸ The equation below can calculate the difference between the possible substituted ions (Ba²⁺, Lu³⁺) and the doped rare ions (Eu²⁺, Nd³⁺) in radius percentage.²⁹

Dr = 100 × [Rm(CN) − Rd(CN)]/Rm(CN) (1)

where Dr is the radius percentage difference; CN is the coordination number; Rm(CN) is the radius of the host cation; and Rd(CN) is the radius of the doped ion. The radii and calculated values of Dr between Eu²⁺/Nd³⁺ and Ba²⁺/Lu³⁺ are shown in Table 1. In light of calculated results from empirical equations, it can be deduced that Nd³⁺ ions could enter into both Ba and Lu sites while Eu²⁺ ions only seem to enter into the Ba sites.

Fig. 1 (a) XRD patterns for Ba_1−2xLu₂Si₃O₁₀:xEu²⁺, xNd³⁺ (x = 0, 0.004, 0.006, 0.008, 0.01 0.02 and 0.03). (b) XRD refinement and the crystal structure of the BaLu₂Si₃O₁₀ host.

Table 1 The radius and Dr values between Eu²⁺/Nd³⁺ and Ba²⁺/Lu³⁺

	CN = 8	CN = 6
	Ba^2+, R = 1.420 Å	Lu^3+, R = 0.861 Å
Eu^2+	R = 1.250 Å, Dr = 12%	R = 1.170 Å, Dr = 35%
Nd^3+	R = 1.109 Å, Dr = 22%	R = 0.983 Å, Dr = 14%

---

3.2 Scanning electron microscopy and transmission electron microscopy analyses

The morphology of the two selected BaLu₂Si₃O₁₀ host and BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ samples was detected using SEM technology. As shown in Fig. 2a and c, the morphologies of these two samples are similar to each other. As shown, the blocky particles are irregularly shaped and approximately 2–10 μm in size, making their form and structure suitable for the fabrication of phosphors. The elemental composition of the BaLu₂Si₃O₁₀ host was investigated by EDX spectroscopy and the spectrum is shown in Fig. 2b. It confirms the presence of Ba, Lu, Si, and O in the BaLu₂Si₃O₁₀ sample. The elements of C and Cu come from the carbon membrane. The distance between the crystal planes reaches 0.325 nm, which is consistent with the (002) planes of BaLu₂Si₃O₁₀ in Fig. 2b (inset).³⁰ In addition, the elemental composition of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ was also investigated by EDX spectroscopy with the spectrum shown in Fig. 2d. As shown in Fig. 2d (inset), the elements of Eu and Nd are also detected.

Fig. 2 (a) SEM images of the BaLu₂Si₃O₁₀ host. (b) EDX spectrum of the BaLu₂Si₃O₁₀ host. (Inset) HRTEM images of the BaLu₂Si₃O₁₀ host. (c) SEM images of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺. (d) EDX spectrum of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺. (Inset) Regional area with high resolution.

3.3 Band structure

Fig. 3a–d show the computations of BaLu₂Si₃O₁₀ abiding by the density functional theory and on the basis of the crystal structure refinement. This paper selects the LDA as the theoretical basis of the density function. The relevant compound takes on an indirect bandgap which reaches approximately 4.843 eV, with the conduction band (CB) peaking at the G point and the valence band (VB) peaking at the D point of the Brillouin zone, which suggests that this is an appropriately wide bandgap for Eu²⁺ to serve as a center to emit light and for Nd³⁺ to serve as a trap center. The electronic structure taken on by the valence band primarily stems from O 2p states, whereas the conduction band is mostly encompassed by Lu 5d states. It is worth mentioning that Ba 5d states also have a tiny contribution to the conduction band. This special band structure characteristic has great effects on the phenomena of photoluminescence and phosphorescence.

Fig. 3 (a) Band structure of BaLu₂Si₃O₁₀. (b) Density of states. (c) Brillouin zone. (d) Total and partial density of states.

3.4 UV-visible diffuse reflectance spectrum

Fig. 4 shows the DRS of the undoped BaLu₂Si₃O₁₀ host and concentration gradient samples of BaLu₂Si₃O₁₀:Eu²⁺. For the host, the absorption in the ultraviolet region ranging from 200 to 400 nm arises from the valence-to-conduction band. For the samples of BaLu₂Si₃O₁₀:Eu²⁺, the feature absorbance peaks of 290 nm and 325 nm are consistent with the excitation peaks as shown in Fig. 5a, which are attributed to the Eu²⁺ ions. The observed white color of the sample is a consequence of the spectral range being in the visible region. Kubelka–Munk-transformed DRS of the BaLu₂Si₃O₁₀ host, in line with the formula of [F(R) × hν]ⁿ = A(hν − E_g), is presented in the inset of Fig. 4 where F(R) is the Kubelka–Munk function with F(R) = (1 − R)²/2R; R is the observed reflectance in the DRS; n = 1/2 for an indirect allowed transition, or 2 for a direct allowed transition, in this case n equals 1/2 according to the above calculated results of the band structure for BaLu₂Si₃O₁₀; A is the proportional constant; ν is the photon energy; and E_g is the value of the host bandgap.²⁹ By adopting the methods proposed by Cao et al.³¹ the optical bandgap energy of BaLu₂Si₃O₁₀ is determined to be 5.5420 eV by extrapolation to F(R) = 0. It is expected that the value of the optical bandgap surpasses that of the previously calculated bandgap of approximately 4.843 eV as the local-density approximation depends less on the bandgap size and spin–orbit coupling of Lu³⁺ which has not been considered.³²

Fig. 4 DRS of the undoped BaLu₂Si₃O₁₀ host and concentration gradient samples of BaLu₂Si₃O₁₀:Eu²⁺. (Inset) A curve of [F(R) × hν]^1/2 vs. hν for BaLu₂Si₃O₁₀.

Fig. 5 (a) Excitation and emission spectra of BaLu₂Si₃O₁₀:0.008Eu²⁺. (Inset) CIE chromaticity diagram and photographs of the BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ phosphor excited at 325 nm. (b) Emission spectra of concentration gradient samples of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺. (Inset) Emission spectra of concentration gradient samples of BaLu₂Si₃O₁₀:Eu²⁺.

3.5 Photoluminescence and phosphorescence properties

The emission and excitation spectra for the BaLu₂Si₃O₁₀:Eu²⁺ phosphor at room temperature is presented in Fig. 5a. The excitation spectrum that was monitored at 426 nm shows a broad excitation band ranging from 250 nm to 400 nm which arises from the 4f⁷–4f⁶5d¹ transition of Eu²⁺ ions. Two primary peaks at 290 nm and 325 nm which are contained in the broad band show that Eu²⁺ ions are present and that their role is as an emission center. In the case of a 325 nm optimal excitation, the material merely shows one wide asymmetric band with a long shoulder peaking at 426 nm. Such an emission peak arises from the Eu²⁺ ions undergoing equally allowed electronic 4f⁶5d¹-4f⁷ transitions, which is considerably affected by the lattice of the host.³⁰ In light of calculated results from the empirical equation, it can be deduced that Eu²⁺ ions only seem to access the Ba sites. However, on the basis of the analysis performed on the emission and excitation spectra, the incorporated Eu²⁺ ions could take up both Lu and Ba sites in the BaLu₂Si₃O₁₀ host. Accordingly, the emission centers fall into two types. In this regard, this work adopts Gaussian deconvolution which takes into account the emission spectrum of the BaLu₂Si₃O₁₀:Eu²⁺ phosphor being widely asymmetric, and the spectrum is compatible with two peaks at 507 nm and 430 nm. As the higher emission intensity shown by Eu1 in contrast to Eu2 elucidates, the Eu²⁺ ions preferably take up Ba sites instead of Lu sites. Fig. 5a presents (0.2018, 0.1595) as the CIE chromaticity coordinate for BaLu₂Si₃O₁₀:Eu²⁺ and blue as the color, which is acquired from the spectrum for emitting light whereby the calculation of the chromaticity coordinate follows the CIE1931 (Commission International de I’Eclairage France) system. As shown in Fig. 5b as the concentration of Nd³⁺ ions increases the emission spectra of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ merely show one band being asymmetrically wide, which is similar to that of the single doped sample. As shown in Fig. 5b (inset), all doped samples have a broad asymmetric emission peak at 426 nm with a long shoulder. It also means Eu²⁺ ions fill both Ba and Lu sites. As the doping concentration increases the brightness increases and then decreases and shows BaLu₂Si₃O₁₀:0.008Eu²⁺ to be the best doped sample. Additionally, this work fails to observe the representative emissions of the Nd³⁺ ions. As the foregoing results indicate, trivalent lanthanide ions Nd³⁺ are not deemed emission centers in BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ but are centers for trapping in numerous LLP phosphors.³⁰

Fig. 6a shows the phosphorescence spectra of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ at different times after removing the excitation source. Interestingly, the profiles do not change with the phosphorescence decay time, but are greatly different from the emission profiles. This interesting phenomenon indicates that the LLP originates from both emissions (Eu_Lu and Eu_Ba) but they have very different contributions during the LLP process. In Fig. 6b, the phosphorescence spectrum of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ which was measured 30 min after the source of excitation was switched off is processed by Gaussian deconvolution and is consistent with two peaks at 430 nm and 507 nm, which is similar to the PL spectrum. The phosphorescence CIE chromaticity coordinate is (0.2062, 0.2614) and the color is bluish-green, as shown in the inset of Fig. 6b. The PL color is blue but the phosphorescence color is bluish-green, indicating that the phosphorescence is derived from the 5d–4f transition of Eu²⁺ ions in the Ba and Lu sites. Eu_Lu has a greater contribution during the LLP process, so the color of phosphorescence tends to be green, rather than blue.

Fig. 6 (a) Phosphorescence spectrum of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺. (b) Phosphorescence spectra of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ 30 min after the removal of the excitation source. (Inset) CIE chromaticity diagram and photograph of the phosphorescence spectrum for the BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ phosphor, excited at 325 nm.

3.6 Luminescence decay curves

To demonstrate that the phosphorescence stems from the Eu_Lu emission in the current BaLu₂Si₃O₁₀ host, luminescence decay curves were measured at an excitation of 325 nm and monitored at 426 and 507 nm. The results are shown in Fig. 7. The luminescence decay curves were able to be adjusted by a double-exponential mode and the average luminescence decay time τ could be ascertained from the formula below.³³

τ = (A₁τ₁² + A₂τ₂²)/(A₁τ₁ + A₂τ₂) (2)

where A₁ and A₂ are constant and τ₁ and τ₂ are the two exponential components of the fitting luminescence decay curve. Table 2 lists the fitting parameters specifically. In line with the equation above and the fitting parameters, the average luminescence decay times monitored at 426 and 507 nm were calculated to be 671 and 2114 ns for the BaLu₂Si₃O₁₀:Eu²⁺ phosphor, and 638 and 1239 ns for the BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ phosphor, respectively. We found that both the luminescence decay times monitored at 426 and 507 nm are characteristic of the 5d–4f allowed transition of Eu²⁺, indicating that the green phosphorescence at 507 nm indeed originates from the Eu²⁺ emission, rather than from some defect levels.³⁴ The Eu²⁺ luminescence decay time decreases with the introduction of Nd³⁺ ions. The reason is that with the introduction of Nd³⁺ ions there is an inequivalence of substitution which leads to lattice distortion and causes lots of defects. Those defects could play a role as energy traps, where the energy from Eu²⁺ ions can transfer. This new approach for energy transfer causes a decrease in the luminescence decay time. Furthermore, the dual-exponential curve fitting arising from the spectral overlap of the two emission bands and can support the statement that Eu²⁺ occupies both the Ba and Lu sites in BaLu₂Si₃O₁₀.³⁰

Fig. 7 Luminescence decay curves of BaLu₂Si₃O₁₀:Eu²⁺ and BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ excited at 325 nm and monitored at 426 nm and 507 nm.

Table 2 The fitting parameters of the luminescence decay curves of BaLu₂Si₃O₁₀:Eu²⁺ and BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ excited at 325 nm and monitored at 426 nm and 507 nm.

Sample	λem (nm)	τ1 (ns)	τ2 (ns)	A1	A2
Eu^2+	426	722.81	235.58	1438.98	524.39
	507	2618.30	470.54	545.66	930.57
Eu^2+, Nd^3+	426	671.06	160.98	1584.67	448.23
	507	1606.05	288.46	313.56	673.59

---

3.7 LLP decay curves

As presented in Fig. 8a, this work measured the LLP decay curves for all of the as-synthesized samples (S1–S6) in order to delve into the decay performance of the BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ emission. The decay process of the LLP is encompassed by a slow decay process and a fast decay process. Those fast decay processes occur initially and dominate the intensity at the very beginning, and the slow decay processes occur later which lead to the long term behavior.^35–38 This work, through adopting curve-fitting technology, analyzes the decay curves of the samples (S1–S6), and these curves are found to adequately match the following double-exponential equation.

I = A₁ exp(−t/τ₁) + A₂ exp(−t/τ₂) + A₀ (3)

where I is the phosphorescence intensity, A₀, A₁ and A₂ are constants, t is time, and τ₁ and τ₂ are the decay times of the exponential components. Since the performance of LLP is chiefly determined by τ₂,³⁹ the τ₂ values for all of the samples are listed in Table 3. The value of τ₂ increases in the beginning and then decreases later with the increase in Eu²⁺ and Nd³⁺ ion content. They reach a maximum when the amount of the Eu²⁺ and Nd³⁺ ions reaches 0.008 mol. LLP decay curves of BaLu₂Si₃O₁₀:Eu²⁺ are shown in the inset of Fig. 8a. The values peak when there is 0.008 mol of Eu²⁺ ions. The initial LLP intensity of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ is in the vicinity of 0.013 cd m⁻² and its LLP can continue in excess of 6 h over an intensity level which is able to be recognized (0.32 mcd m⁻²). The phosphorescence of BaLu₂Si₃O₁₀:0.008Eu²⁺ is only visible for 199 s by the naked eye, which indicates that the introduction of Nd³⁺ ions greatly optimizes the phosphorescence of BaLu₂Si₃O₁₀:0.008Eu²⁺. The LLP decay curve of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ is also plotted as a function of reciprocal persistent luminescence intensity (I⁻¹) versus time (t), as presented in Fig. 8b. The detailed reasons are discussed later in the section on TL curves.

Fig. 8 (a) LLP decay curves of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺. Inset: LLP decay curves of BaLu₂Si₃O₁₀:Eu²⁺ (b) function of reciprocal LLP intensity (I⁻¹) versus time (t) of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ excited for 10 min.

Table 3 Decay times for two exponential components of Ba_1−2xLu₂Si₃O₁₀:xEu²⁺, xNd³⁺ (x = 0, 0.004, 0.006, 0.008, 0.01 0.02 and 0.03)

Sample	A0	A1	τ1 (s)	A2	τ2 (s)
S1	7.2266 × 10^−4	0.0110	55.5423	0.0054	1224.0458
S2	6.7820 × 10^−4	0.0086	62.6299	0.0046	1242.8820
S3	8.3097 × 10^−4	0.0170	60.0109	0.0065	1326.3692
S4	6.6350 × 10^−4	0.0093	50.3919	0.0037	1323.1464
S5	8.0447 × 10^−4	0.0150	55.4197	0.0061	1191.9548
S6	3.9931 × 10^−4	0.0022	26.1850	0.0012	236.5758

---

3.8 Thermoluminescence curves

Trapping centers play a momentous role in the photo-energy storage of persistent and thermo-stimulated phosphors.⁴⁰ After removal of the excitation source, the captured charge carriers are able to escape due to thermal disturbance and are transferred to centers for luminescence, which triggers the phosphorescence. Accordingly, shallow depth traps are negatively and adversely affected by stabilizing the charge carriers, which considerably shortens the duration period of persistent luminescence. In contrast, the charge carriers that are captured by deep depth traps are extremely hard to release at room temperature, which leads to the defective performance of the LLP.^40–42 For this reason, as the appropriate trap depth is formed, a solid foundation is laid for attaining a perfect LLP performance. Meanwhile, the trap density is also a significant factor for LLP. To characterize the traps in our samples TL measurements were conducted on BaLu₂Si₃O₁₀:0.008 Eu²⁺ and BaLu₂Si₃O₁₀:0.008 Eu²⁺, 0.008 Nd³⁺, their TL curves are shown in Fig. 9, we only observe three very weak peaks in BaLu₂Si₃O₁₀:Eu²⁺ and it is possible that the TL bands arise from intrinsic defects in the host as the Eu²⁺ ions are equally replaced by Ba²⁺ ions in BaLu₂Si₃O₁₀. Additionally, as the weak TL signal indicates, the concentration of carriers that is captured by the intrinsic defects is extremely low. When Nd³⁺ ions are co-doped, the intensity of the peaks are evidently enhanced, and the primary peak position moves to 330 K, meaning that the doping of the Nd³⁺ ions considerably elevates the defect levels. To further evaluate the trap states, the trap depths (E_t) and the trap densities (n₀) can be acquired from eqn (4) and (5).⁴³

E_t = [2.52 + 10.2 × (μ_g − 0.42)] × (K_BT_m²/ω) − 2K_BT_m (4)

n₀ = ωI_m/{β × [2.52 + 10.2 × (μ_g − 0.42)]} (5)

Where I_m is the intensity of the TL peak, δ is the high-temperature half-width, τ is the low-temperature half-width, the asymmetry parameter μ_g = δ/(τ + δ), β is the heating rate, K_B is the Boltzmann constant (1.38 × 10⁻²³ J K⁻¹), T_m is the temperature of the TL peak, and ω is known as the shape parameter and is defined as ω = τ + δ. As shown in Table 4, for the BaLu₂Si₃O₁₀:0.008Eu²⁺ sample, although the trap depth is appropriate for releasing the trapped carriers, the low trap density directly results in a brief LLP of 199 s. For BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺, by co-doping with Nd³⁺ ions, more traps are generated with a suitable energy depth which greatly improves the duration of the LLP and the LLP brightness in the very beginning. This result suggests that co-doping indeed has a great effect upon the LLP performance of the phosphor.

Fig. 9 TL curves of BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺ excited for 25 min. Inset: concentration gradient TL curves of BaLu₂Si₃O₁₀:Eu²⁺, Nd³⁺ excited for 5 min.

Table 4 TL parameters of BaLu₂Si₃O₁₀:0.008Eu²⁺ and BaLu₂Si₃O₁₀:0.008Eu²⁺, 0.008Nd³⁺

Sample	Glow-peak parameters					TL parameters
	Tm (K)	T	δ	ω	μg	Et (eV)	n0
Eu^2+	319	19	23	42	0.55	0.75	2.65 × 10^5
Eu^2+, Nd^3+	330	21	27	48	0.56	0.71	3.75 × 10^6

---

3.9 LLP mechanism

Under normal conditions, for the majority of LLP materials, the color of the phosphorescence is identical to the color of the PL arising from the identical emission centers. Yet the blue emitting phosphors show a phosphorescence which is characterized as bluish-green in this paper. In line with the previous analysis, Eu²⁺ ions which occupy Ba and Lu sites serve as emission centers and Nd³⁺ ions which take up Ba sites serve as trapping centers in the co-doped samples. As these experiments overall indicate, this material provides highly-efficient trapping conditions. A probable LLP mechanism for forming the efficient bluish-green LLP in Eu²⁺ and Nd³⁺ co-doped BaLu₂Si₃O₁₀ is shown in Fig. 10. Under UV excitation, the 4f⁷ ground-state electrons of the Eu²⁺ ions are promoted to the 4f⁶5d¹ excited state (excitation process 1). Partial electrons which are promptly excited are reincorporated with the Eu²⁺ ions excited on the Ba and Lu sites to produce luminescence (PL process 2). As most Eu²⁺ ions occupy the Ba sites and tiny Eu²⁺ ions occupy the Lu sites, the phosphor emits blue light. Simultaneously other excited electrons are able to move via the conduction band and are captured later by the traps (electron trapping process 3). When the electrons captured by shallow traps are thermally released at room temperature, they move back to the conduction band directly and a few of the captured electrons in deep traps escape through a quantum tunneling process. In view of the conduction band primarily being formed by Lu 5d states rather than Ba states, the released electrons prefer to transfer to the excited Eu²⁺ ions on the Lu sites which are in the same form of resonance and recombine with them, which gives rise to the green phosphorescence (LLP process 4). In short, it is the special form of the conduction band and the occupation situation of the emission centers that result in the interesting optical phenomenon.

Fig. 10 A possible schematic diagram of the afterglow mechanism.

4. Conclusions

In a nutshell, a newly proposed bluish-green emitting long-lasting phosphor with prominent LLP properties is successfully achieved via a high temperature solid state method. Eu²⁺ ions show a broad asymmetric emission band with a peak at 426 nm as it occupies both the Ba and Lu sites in BaLu₂Si₃O₁₀. The blue emitting phosphors show a kind of green phosphorescence phenomenon, which is induced mostly by the emission of Eu²⁺ ions occupying the Lu sites. Co-doping with Nd³⁺ ions can effectively improve the LLP performance. The LLP can last more than 6 h above the recognizable intensity level (0.32 mcd m⁻²). This work also provides direct experimental evidence that the conduction band is of critical significance for the course of LLP. And the analysis regarding this experimental work backs up the LLP mechanism. This new LLP material exhibits a 6 h LLP and has considerable potential for extensive applications.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work is supported by the National Natural Science Funds of China (Grant No. 51672115 and Grant No. 51372105). Thanks for the support of Chengguan district Lanzhou city science and technology development projects (project number: 2017-2-2).

Notes and references

1. Z. Pan, Y. Y. Lu and F. Liu, Nat. Mater., 2012, 11, 58–63.
2. C. N. Xu, T. Watanabe, M. Akiyama and X. G. Zheng, Appl. Phys. Lett., 1999, 74, 2414–2416.
3. D. Q. Chen, Y. Chen, H. W. Lu and Z. G. Ji, Inorg. Chem., 2014, 53, 8638–8645.
4. D. Q. Chen, J. Eur. Ceram. Soc., 2014, 34, 4069–4075.
5. T. Maldiney, C. Richard, D. Scherman and D. Gourier, Chem. Mater., 2014, 26, 1365–1373.
6. J. Wang, Q. Su and S. B. Wang, J. Phys. Chem. Solids, 2005, 66, 1171–1176.
7. S. Lian, Y. Qi, C. Rong, L. Yu, A. Zhu, D. Yin and S. Liu, J. Phys. Chem. C, 2010, 114, 7196–7204.
8. B. Wang, H. Lin, Y. L. Yu, D. Q. Chen, R. Zhang, J. Xu and Y. S. Wang, J. Am. Ceram. Soc., 2014, 97, 2539–2545.
9. T. Matsuzawa, Y. Aoki, N. Takeuchi and Y. Murayama, J. Electrochem. Soc., 1996, 143, 2670–2673.
10. H. J. Guo, W. B. Chen, W. Zeng and Y. H. Wang, J. Mater. Chem. C, 2015, 22, 5844–5850.
11. Z. Hong, P. Zhang, X. Fan and M. Wang, J. Lumin., 2007, 124, 127–132.
12. V. K. Van den Eeckhout, P. F. Smet and D. Poelman, J. Lumin., 2009, 129, 1140–1143.
13. T. Matsuzawa, Y. Aoki, N. Takeuchi and Y. Murayama, J. Electrochem. Soc., 1996, 143, 2670–2673.
14. H. Yamamoto and T. Matsuzawa, J. Lumin., 1997, 72–74, 287–289.
15. I. P. Sahu, D. P. Bisen, N. Brahme, L. Wanjari and R. K. Tamrakar, Res. Chem. Intermed., 2015, 41, 8797–8814.
16. P. J. Wang, X. H. Xu, D. C. Zhou, X. Yu and J. B. Qiu, Inorg. Chem., 2015, 54, 1690–1697.
17. T. L. Barry, J. Electrochem. Soc., 1968, 115, 1181–1183.
18. Y. Lin, Z. Zhang, Z. Tang, X. Wang and Z. Zheng, J. Eur. Ceram. Soc., 2001, 21, 683–685.
19. M. Wierzbicka-Wieczorek, U. Kolitsch and E. Tillmanns, Eur. J. Mineral., 2010, 22, 245.
20. Q. Wang, G. Zhu, S. Y. Xin, X. Ding, J. Xu, Y. S. Wang and Y. H. Wang, Phys. Chem. Chem. Phys., 2015, 17, 27292–27299.
21. J. Q. Wang, W. R. Zhao, J. M. Zhong and L. C. Lan, J. Mater. Sci.: Mater. Electron., 2014, 25, 2162–2168.
22. J. Zhou and Z. G. Xia, J. Mater. Chem. C, 2015, 3, 7552–7560.
23. W. R. Liu, C. C. Lin, Y. C. Chiu, Y. T. Yeh, S. M. Jang, R. S. Liu and B. M. Cheng, Opt. Express, 2009, 17, 18103–18109.
24. J. Zhou and Z. G. Xia, Opt. Mater., 2016, 52, 116–122.
25. J. Brgoch, K. Hasz, K. A. Denault, C. K. H. Borg, A. A. Mikhailovsky and R. Seshadri, Faraday Discuss., 2014, 176, 333–347.
26. K. Li, S. S. Liang, H. Z. Lian, M. M. Shang, B. G. Xing and J. Lin, J. Mater. Chem. C, 2016, 4, 3443–3453.
27. P. Dorenbos, J. Phys.: Condens. Matter, 2003, 15, 8417.
28. M. Y. Peng, Z. W. Pei, G. Y. Hong and Q. Su, J. Mater. Chem., 2003, 13, 1202–1205.
29. R. D. Shannon, Acta Crystallogr., 1976, 32, 751–767.
30. G. Li, Y. H. Wang, H. J. Guo, J. Liu, D. W. Liu and P. Feng, J. Lumin., 2017, 192, 98–104.
31. G. Cao, L. K. Rabenberg, C. M. Nunn and T. E. Mallouk, Chem. Mater., 1991, 3, 149–156.
32. Y. Q. Li, Y. H. Wang, X. H. Xu, G. Yu and F. Zhang, J. Electrochem. Soc., 2010, 157, 39–43.
33. N. Ruelle, M. P. Thi and C. Fouassier, Jpn. J. Appl. Phys., 1992, 31, 2786–2790.
34. C. Y. Liu, Z. G. Xia, Z. P. Lian, J. Zhou and Q. F. Yan, J. Mater. Chem. C, 2013, 1, 7139–7147.
35. J. Zhang, X. Ma, Q. Qin, L. Shi, J. Sun, M. Zhou, B. Liu and Y. Wang, Mater. Chem. Phys., 2012, 136, 320–324.
36. J. Wang, S. Wang and Q. Su, J. Mater. Chem., 2004, 14, 2569–2574.
37. R. Pang, C. Li, L. Shi and Q. Su, J. Phys. Chem. Solids, 2009, 70, 303–306.
38. B. Lei, H. Zhang, W. Mai, S. Yue, Y. Liu and S. Man, Solid State Sci., 2011, 13, 525–528.
39. T. Wang, Y. H. Hu, L. Chen, X. J. Wang and G. F. Ju, Radiat. Meas., 2015, 73, 7–13.
40. J. Glodo and A. Wojtowicz, J. Alloys Compd., 2000, 300, 289–294.
41. T. Aitasalo, P. Deren, J. Hölsä, H. Junger, J. C. Krupa, M. Lastusaari, J. Legendziewicz, J. Niittykoski and W. Strek, J. Solid State Chem., 2003, 171, 114–122.
42. T. Kinoshita, M. Yamazaki, H. Kawazoe and H. Hosono, J. Appl. Phys., 1999, 86, 3729–3733.
43. R. Chen, J. Appl. Phys., 1969, 40, 570–585.

---