2.1.

Synthesis of Phosphors

A series of ${\mathrm{LiTaO}}_3:x{\mathrm{Tb}}^{3+}$ ($x=0$, 0.5%, 1%, 2%) phosphors were synthesized by high-temperature solid-state reaction. The starting materials ${\mathrm{Li}}_2{\mathrm{CO}}_3$ (4N), ${\mathrm{Ta}}_2{\mathrm{O}}_5$ (4N), and activator ${\mathrm{Tb}}_4{\mathrm{O}}_7$ (4N) were used without further purification. The raw materials of phosphors activated by other elements (e.g., Eu) were also prepared using the corresponding oxides, such as ${\mathrm{Eu}}_2{\mathrm{O}}_3$ (4N). First, the stoichiometric amounts of raw materials were ground in a mortar with an appropriate amount of anhydrous ethanol for 30 min until the complete evaporation of ethanol. Then, the mixture was pre-sintered for 2 h at 600°C in a muffle furnace, cooled to RT, and fully ground. It was finally sintered at 1100°C for 5 h in air to obtain the phosphors.

2.2.

Material Characterization

X-ray diffraction (XRD) patterns of the as-prepared phosphors were recorded on a Bruker D8 ADVANCE X-ray diffractometer with Cu K radiation ($\lambda =0.15418\mathrm{nm}$). The composition and morphology of the powders were investigated on a JSM-6700F field emission scanning electron microscope as well as a JEM-2100F transmission electron microscope (Japan JEOL Corporation). X-ray absorption fine structure (XAFS) of Ta and Tb ${\mathrm{L}}_3$ edge was obtained at Taiwan Light Source (BL17C1). XAFS data were processed using the Demeter XAS data processing package.²⁴ For the Ta ${\mathrm{L}}_3$ edge, the data were fit in R-space, from 1 to 2.4 Å. Two Ta-O paths of varied lengths were used in the fitting. The R-space extended X-ray absorption fine structure (EXAFS) signal was obtained by a variable $k^n$-weighted Fourier transform ($n=1,2,3$) of the EXAFS signal $\chi (k)$ over a $k$-range of 3 to $11.5{\AA }^{-1}$. Photoluminescence (PL) and photoluminescence excitation (PLE) spectra were recorded by a Hitachi F-4600 fluorescence spectrophotometer with a 150 W xenon arc lamp as the light source. The temperature-dependent PL spectra were measured with TAP-02 high-temperature fluorescence test accessories. TL spectra were conducted by an SL008 TL meter (Radiation Science and Technology Co. Ltd, Guangzhou, China). UV irradiation for 5 min was used to charge the phosphor particles. Until the afterglow dissipated, the sample was heated to a certain temperature at a rate of 1 K/s to record the signals. X-ray photoelectron spectroscopy (XPS) measurements were performed by electron spectroscopy for a chemical analysis system (Axis Ultra, Kratos analytical-A Shimadzu Group company) with an Al (Mono) source (150 W). In situ temperature-dependent Raman characterization was performed using a Renishaw inVia Laser confocal Raman Spectrometer. The laser wavelength was 523 nm. Raman spectra were collected in the temperature range from 305 to 498 K.

In a PALS experiment, a ${}^{22}\mathrm{Na}$ radioactive source was used as the positron source with a source intensity of about $13\mu \mathrm{J}$. Two identical samples were clamped on either side of a ${}^{22}\mathrm{Na}$ source to form a typical sandwich geometry. The conventional positron annihilation lifetime spectrometer uses a pair of ${\mathrm{BaF}}_2$ scintillators to detect gamma rays released after positron annihilation. The positron annihilation lifetime spectrum was collected by the fast-slow coincidence measurement technology. The total lifetime spectrum number is 2 million to ensure the validity of the statistics. The time resolution of the spectrometer was 195 ps, and the electronic plug-in of the measurement system was the standard NIM plug-in of EG&G Company.

The performance of pressure sensing was studied by a laboratory-built integrated test system using a computer-controlled Shimadzu Universal stress testing instrument (AGS-X 10KN STD) with a photomultiplier tube (Hamamatsu photon, C13796) and a fluorescence spectrometer (SR-500I-B1-R, Andor).

2.3.

Synthesis of Optical Information Storage Wafer

A certain amount of polydiethylsiloxane (PDMS) precursor and curing agent was put into a beaker, followed by adding phosphor particles with a phosphor/PDMS weight ratio of 0.4:1. After stirring evenly, the mixture was slowly poured into a 4 cm × 4 cm mold. A flexible wafer was obtained by heating and curing at 75°C for 1 h in an oven.

2.4.

Preparation of Pressure Sensing Devices

The powder sample of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ was mixed with optical epoxy resin, followed by injection into a cylindrical mold container with a diameter of 20 mm and a height of 15 mm. After curing at 50°C for 6 h, the samples were taken from the mold. Before testing, the cured sample was exposed to UV radiation at 254 nm for 5 min.

2.5.

DFT Theoretical Calculation

Theoretical calculation of the electronic structure was performed based on the density functional theory (DFT) implemented via the Vienna ab initio simulation package (VASP) code.²⁵ Moreover, the exchange correlation potential is selected by the generalized gradient approximation (GGA) with the Perdew–Burke–Ernzerhof (PBE) formulation. The cutoff energy was 400 eV, and Brillouin zone integration was represented using the $K$-point sampling scheme of a $4\times 4\times 2$ Monkhorst–Pack grid. The stopping criterion for ionic optimization number is $-0.03{\AA }^{-1}$, while the criterion for allowed error in total energy is $1.0\times {10}^{-6}\mathrm{eV}$. The formation energies ($E_{\text{for}}$) of ${\mathrm{LiTaO}}_3$ with different defects were calculated by²⁶

2.6.

Calculation of Positron Annihilation Lifetime

Positron annihilation lifetime calculations were performed using an ABINIT software with two-component density-functional theory (TCDFT) calculations in the projector augmented-wave (PAW) framework. Calculations were performed using a $2\times 2\times 1$ ${\mathrm{LiTaO}}_3$ supercell consisting of 120 atoms. In the defective structural models, a Ta vacancy (${\mathrm{V}}_{\mathrm{Ta}}^{5-}$) was built by removing a Ta atom from the supercell, and a Tb atom was further applied to substitute a Li atom to form a complex-defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$.

3.1.

Structural Characterizations

The XRD patterns illustrate that all the as-prepared samples correspond well with standard ${\mathrm{LiTaO}}_3$ hexagonal phase (PDF#29-0836) [Fig. 2(a) and Fig. S1 (see Supplementary Material)]. As shown in Fig. 2(b), the XRD Rietveld refinement of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ implies that the calculated data comply with the experimental measurements, where each Ta atom and Li atom are coordinated with six O atoms affording Ta-O and Li-O octahedron units (Table S1 in the Supplementary Material).^27–29 Moreover, the energy-dispersive X-ray spectroscopy (EDS) elemental mapping images show the uniform distribution of Ta, O, and Tb elements throughout phosphor particles of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ [Fig. 2(c)].

Fig. 2

Structural characterizations of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ phosphor. (a) XRD patterns for ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ and ${\mathrm{LiTaO}}_3$. (b) XRD Rietveld refinement of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ with the measured and calculated data. (c) SEM and EDS mapping images of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$, implying the uniform doping of Tb throughout the whole particle. (d), and (e) HRTEM images of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$. (f) SAED (i) and filtered images (ii–iv) of regions within dashed white squares in (d)–(e). (g) Intensity profiles along the $\mathrm{L}1$–$\mathrm{L}5$ lines recorded from (f) (ii–iv). The results confirm the existence of the ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ defects in the lattice of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$. (h)–(j) DFT calculations of ${\mathrm{LiTaO}}_3$ with and without ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$,${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ defects, where gray, light blue, white, and purple spheres represent Li, Ta, O, and Tb atoms, respectively. In (h), the solid line represents the total DOS for perfect lattice ${\mathrm{LiTaO}}_3$, while the dotted line represents the total DOS for ${\mathrm{LiTaO}}_3$ with ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$.

The HRTEM images of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ indicate that the lattice fringe with $d$ spacing of 0.372 nm corresponding to the (012) plane [Figs. 2(d)–2(e)] is well consistent with the refined structure of hexagonal phase ${\mathrm{LiTaO}}_3$. Also, the selected area electron diffraction (SAED) pattern taken along the axis of [$\overline{221}$] zone confirms the single-crystal nature of the phosphor particles [Fig. 2(f) panel (i)].³⁰ Figure 2(f) panels (ii)–(iv) show the filtered images of the regions within dashed white squares in Figs. 2(d) and 2(e). The atomic arrangement of Ta atoms observed from the images, as well as the recorded intensity profile verifies the phase of ${\mathrm{LiTaO}}_3$ [$\mathrm{L}1$–$\mathrm{L}5$ in Fig. 2(f)], consistent with the XRD refinement result. The filtered images display that the lattice fringes are discontinuous, and some lattice positions are blurred or even missing, indicating the existence of defects [Fig. 2(f) panels (iii) and (iv) and Figs. S2, S3 in the Supplementary Material].³¹ In contrast to the intensity profiles of perfect lattice $\mathrm{L}1$, and $\mathrm{L}2/\mathrm{L}3$ with significant absence of Ta atoms (${\mathrm{V}}_{\mathrm{Ta}}^{5-}$), Tb can be observed to occupy the position of Li atoms between the Ta atomic gap, forming the aliovalent substitution defects ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, as shown in $\mathrm{L}4$. Furthermore, since the ionic radius of ${\mathrm{Li}}^{+}$ ($r=0.76\AA$, $\mathrm{C}\mathrm{N}=6$) is larger than that of ${\mathrm{Ta}}^{5+}$ ($r=0.64\AA$, $\mathrm{C}\mathrm{N}=6$), the doped ${\mathrm{Tb}}^{3+}$ ($r=0.92\AA$, $\mathrm{C}\mathrm{N}=6$) with a large radius is inclined to occupy the ${\mathrm{L}\mathrm{i}}^{+}$ lattice site. Also, the intensity distribution recorded along $L5$ verifies the existence of the complex defects $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ composed of adjacent ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$. These defects ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ are expected as capture sites with different depths for storing carriers, compensating for the TQ effect, and thus increasing the emission intensity of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ under high temperature.^7,32

The formation of triple defects in ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ was also investigated from the thermodynamic point of view by first-principles calculations. The formation energies ($E_{\text{for}}$) of possible defects, including ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, Li vacancy (${\mathrm{V}}_{\mathrm{Li}}^{-}$), inequivalence defects ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ and ${\mathrm{Tb}}_{\mathrm{Ta}}^{2-}$, as well as complex defects $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ and $({\mathrm{V}}_{\mathrm{Li}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{+}$, are calculated as shown in Fig. S4 in the Supplementary Material. The results show that ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and complex defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ have lower formation energy than ${\mathrm{V}}_{\mathrm{Li}}^{-}$, ${\mathrm{Tb}}_{\mathrm{Ta}}^{2-}$, and $({\mathrm{V}}_{\mathrm{Li}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{+}$, also verifying the existence of triple defects ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ as observed by HRTEM. Furthermore, the calculated energy band structures show that these defects have no significant effect on the bandgap (Fig. S5 in the Supplementary Material), but impact the electronic structure greatly. Density of states (DOS) of ${\mathrm{LiTaO}}_3$ with these defects provides us with the specific effects of defects on the electronic structure [Figs. 2(h)–2(j)]. By comparing the total DOS profiles, ${\mathrm{LiTaO}}_3$ with ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ displays a higher occupied state at the bottom of the valence band (VB), leading to a better carrier storage capacity. It is evident that ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ can form a new intermediate level near the conduction band (CB). As an energy buffer layer, it is mainly composed of ${\mathrm{Tb}}^{3+}4\mathrm{f}$ orbitals, which can realize the ET from the defect level to the ${\mathrm{Tb}}^{3+}$ luminous level. Also, the presence of the complex defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ has the same effect. A new intermediate level mainly consisting of ${\mathrm{Tb}}^{3+}4\mathrm{f}$ orbitals appears near the VB. This indicates that the complex defect can effectively improve the $\mathrm{f}$-electron domination near the Fermi level, associated with the promoted carrier transfer.³³ All these characterizations confirm the existence of triple defects with varied electronic structures in ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$, which could act as the energy buffer layers to promote NTQ.

3.2.

NTQ Behaviors and Defect Identification of LiTaO₃:Tb³⁺ Phosphor

To have a closer insight into the effect of defects on NTQ, TL, as one of the most effective techniques for characterizing the type and depth of defects,^34,35 was measured for the ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ phosphor. The result demonstrates three successive temperature ranges of 300 to 373 K (${\mathrm{T}}_1$), 373 to 423 K (${\mathrm{T}}_2$), and 423 to 498 K (${\mathrm{T}}_3$), indicating three significant trap levels induced by triple defects associated with ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$, respectively [Fig. 3(a) and Figs. S6, S7 in the Supplementary Material]. This result is consistent with the aforementioned computational defect formation energy for these triple defects. The trap depths of defects were calculated using the following empirical equation³⁶:

Fig. 3

Identification of defects in ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$. (a) TL curve of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ with the fitted three peaks ${\mathrm{T}}_1,{\mathrm{T}}_2$, and ${\mathrm{T}}_3$ in different temperature zones, corresponding to three defects, respectively. (b) The calculated trap depths and integral areas of three defects [${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$]. (c) Experimental positron annihilation lifetime spectrum of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ with deconvolution fitting components. The results show two components, ${\tau }_1$ and ${\tau }_2$, ascribed to ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$, respectively. The inset shows the percentages of the ${\tau }_1$ and ${\tau }_2$ components. (d) The relative ratios of two defects [${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$] in ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ with ${\mathrm{Tb}}^{3+}$ doping concentrations, obtained from the TL and PALS deconvolution results, respectively. (e) and (f) Positron density isosurface localized at ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ defects obtained using ABINIT. Yellow represents the positron accumulation. (g) TL curves of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ before and after reduction treatment at 1273 K for 2 h in mixed 95% Ar/5% H₂ atmosphere. (h) Comparison of thermal stability of different LiTaO₃:Tb³⁺ samples. (i) PL decay curves of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ at increasing temperature from 305 to 498 K. These results provide sufficient evidence that triple defects with varied trapping depths in ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ could promote NTQ performance in a wide temperature range.

To better elucidate how defects improve the NTQ performance of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$, the PALS combined with TCDFT calculation was performed to study the defect positron states, enabling the process of trapping and detrapping to be unambiguously established.^37–39 PALS is an effective means of identifying neutral or negatively charged vacancy-related defects in materials because it depends on the local structure of the defects enriched with positrons.^40,41 The experimental positron annihilation lifetime spectrum of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ was deconvolved to determine the components contributing to the positron state [Fig. 3(c)]. There are two dominant defect components ${\tau }_1$ and ${\tau }_2$ with lifetimes of 173.6 and 307.9 ps, contributing to the percentages of 54.7% and 42.4%, respectively. Moreover, the PALS of all samples with different Tb doping concentrations shows two similar deconvolved components (Fig. S9 and Table S2 in the Supplementary Material). The longest lifetime ${\tau }_0$ is attributed to the pick-off annihilations of the ortho-positronium (o-Ps) in the free volume sites, present mainly in the amorphous regions of the polymer matrix. And it is not the main component obtained from the fit, occupying a very small fraction of 0.7% to 1.8%.

The corresponding positron state, including positron lifetime and density, was calculated using TCDFT with ABINIT, as shown in Table S3 in the Supplementary Material and Figs. 3(e) and 3(f). The calculated lifetime of LiTaO₃ with ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ is 173.8 ps, which is close to the ${\tau }_1$ component (173.6 ps) of the experimental results. Meanwhile, the$({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ lifetime is 299 ps. Thus, it is evident that the main positron lifetime components ${\tau }_1$ and ${\tau }_2$ can be attributed to ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and complex defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$, respectively. In addition, the ${\tau }_1$ value is also consistent with the reported positron lifetime range (170 to 200 ps) of B-site vacancy in perovskite oxides ABO₃.^42–44 The relative percentage of two defects ${\mathrm{V}}_{\mathrm{Ta}}^{5-}({\mathrm{T}}_1)$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ (T₃) obtained by TL spectra has the same trend as that obtained by PALS [Fig. 3(d)], validating the existence of these two defects. Figures 3(e) and 3(f) clearly illustrate the positron density at the defects of ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$. It is notable that the positron density in a perfect LiTaO₃ lattice is an order of magnitude lower than that in defects, as illustrated in Fig. S10 in the Supplementary Material. The high positron density caused by defects in the vicinity of the activator indicates that the negative charge defects, ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$, have a strong ability to trap carriers, which could provide a source for ET from the defects to the 4f levels of Tb³⁺.

To demonstrate the important role of triple defects in LiTaO₃:1%Tb³⁺ in promoting NTQ, a series of samples with regulatory defects were designed and prepared for comparison, including LiTaO₃:1%${\mathrm{Tb}}^{3+}$ with reductive atmosphere treatment ($\mathrm{R}\text{-}{\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$), Li or Ta excessive ${\mathrm{Li}}_{1.005}{\mathrm{TaO}}_3:1\%{\mathrm{Tb}}^{3+}$ or ${\mathrm{LiTa}}_{1.005}{\mathrm{O}}_3:1\%{\mathrm{Tb}}^{3+}$, and ${\mathrm{Eu}}^{3+}$, ${\mathrm{Tb}}^{3+}$ co-doped ${\mathrm{LiTaO}}_3$ phosphor. The TL curve of R-${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ shows that the density of defects in ${\mathrm{T}}_1$ remains unchanged, while those in ${\mathrm{T}}_2$ and ${\mathrm{T}}_3$ evidently decrease [Fig. 3(g)]. As expected, it shows weak NTQ in the low temperature range of 300 to 373 K, and TQ of PL emission can be detected as the temperature increases from 373 to 473 K [Fig. 3(h) and Fig. S11 in the Supplementary Material]. This is because the reducing atmosphere treatment increases the concentration of positively charged oxygen vacancy competing with the ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ defect, thus eliminating the effects of ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ and complex defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$. For the samples with excessive Li or Ta alone [Fig. 3(h)], the defect content in the samples decreased, thus leading to significant TQ.

By comparing Eu³⁺ single-doped and Eu³⁺, Tb³⁺ co-doped LiTaO₃ phosphors (Fig. S12 in the Supplementary Material), the TL results imply that co-doping Tb significantly enhances the thermal stability of phosphors ascribed to rich defects in the co-doped materials. This further certifies that the defects play an important role in suppressing TQ. In agreement with previous reports,⁴⁵ upon increasing the temperature to 498 K, the PL dynamics in ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ is accelerated by $\sim 35\%$ [effective lifetime from $\tau (305\mathrm{K})=32.3\mathrm{ms}$ to $\tau (373\mathrm{K})=21.3\mathrm{ms}$, Fig. 3(i)], suggesting that carrier trapping is strongly enhanced at high temperatures. The improved carrier separation and transfer efficiency were also verified by photoelectrochemical experiments. Figure S13 in the Supplementary Material illustrates that ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ has a larger current density and a smaller charge transfer resistance than ${\mathrm{LiTaO}}_3$. It shows the photogenerated electron-hole separation under UV radiation is better for ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$.

3.3.

Chemical Environments and Structures

Except for the defects effect, the chemical environments and intrinsic structures of ${\mathrm{LiTaO}}_3$ and ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ were also evaluated. In the XPS spectra of Ta $4\mathrm{f}$ [Fig. 4(a) and Fig. S14 in the Supplementary Material], ${\mathrm{LiTaO}}_3$ shows the Ta $4{\mathrm{f}}_{7/2}$ and $4{\mathrm{f}}_{5/2}$ peaks at 26.2 and 28.3 eV, respectively, confirming the dominant oxidation state of ${\mathrm{Ta}}^{5+}$, whereas for ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$, these two peaks are shifted toward lower binding energy, which implies that the as-formed ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ leads to a lower valence state.⁴⁶ In the Li 1s spectra (Fig. S15 in the Supplementary Material), ${\mathrm{LiTaO}}_3$ exhibits the peaks locating at 55.16 eV. For ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$, the peak position shifts toward lower binding energy of 54.92 eV. This results from partial occupation of the Li sites by low concentration of ${\mathrm{Tb}}^{3+}$. Raman spectroscopy is an important tool for studying phonon patterns and assessing local structures in the NTQ materials.^3,47 As shown in Fig. 4(b), ${\mathrm{\nu }}_1$ and ${\mathrm{\nu }}_2$ represent the modes of Ta-O framework deformation.⁴⁸ The ratio of $I_{{\mathrm{\nu }}_1}/I_{{\mathrm{\nu }}_2}$ increases from 1.1 to 1.5 when doping Tb³⁺, indicating that the defects induce the change of the Ta-O deformation.⁴⁹

Fig. 4

Chemical environments and structures of ${\mathrm{LiTaO}}_3$ and ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$. (a) High-resolution Ta $4\mathrm{f}$ XPS spectra and (b) Raman spectra of LiTaO₃:1%Tb³⁺ and ${\mathrm{LiTaO}}_3$. ${\mathrm{\nu }}_1$ and ${\mathrm{\nu }}_2$ stand for the modes of Ta-O framework deformation. The inset in (b) shows the integrated area ratios of ${\mathrm{\nu }}_1/{\mathrm{\nu }}_2$. (c) (i) Temperature-dependent in situ Raman spectra with the temperature range from 305 to 498 K, and (ii) Raman shifts of different modes with changing temperatures. ${\mathrm{\nu }}_3$, ${\mathrm{\nu }}_4$, and ${\mathrm{\nu }}_5$ represent Li–O stretching, O-Ta-O bending mode, and Ta-O stretching mode, respectively. (d) The EXAFS curves in R-space of ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ and ${\mathrm{LiTaO}}_3$. (e) XANES spectra of the Tb L3 edge, recorded from ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ and ${\mathrm{Tb}}_4{\mathrm{O}}_7$ reference sample.

In situ temperature-dependent Raman spectra are shown in Fig. 4(c). With the increase of temperature from 305 to 498 K, all Raman peaks become weaker in intensity and wider in shape, which is associated with the improvement of disordered or relaxation characteristics. There is a significant blueshift of the ${\mathrm{\nu }}_1$, ${\mathrm{\nu }}_3$, and ${\mathrm{\nu }}_5$ vibrational modes, providing a favorable pathway for thermal activation enhancement. The O-Ta-O bending mode at $462{\mathrm{cm}}^{-1}$ (${\mathrm{\nu }}_4$) is strongly coupled with Li-O stretching (${\mathrm{\nu }}_3$) and Ta-O stretching (${\mathrm{\nu }}_5$). The O-Ta-O bending (${\mathrm{\nu }}_4$) is almost constant with increasing temperature, suggesting that it could act as a structural buffer layer for thermally induced fluctuations to reduce thermal effects associated with the structural relaxation of the Tb³⁺ luminescent center. As the temperature increases, the phonon energy increases, and the ET from the defects to the luminescence energy level makes the lattice stable enough to suppress TQ. In addition, the reduction of the maximum phonon energy ($662{\mathrm{cm}}^{-1}$) may favor the luminescence of Tb³⁺ at high temperatures because the nonradiative relaxation could be effectively inhibited.⁴⁷

The EXAFS data as well as the fitting results are shown in Fig. 4(d) and Fig. S16 and Table S4 in the Supplementary Material. The main peak at $\approx 1.96\AA$ is attributed to the scattering interaction between Ta and O atoms (Ta-O). And the average bond length of Ta-O increases from 1.967 to 1.973 Å with the increase of Tb doping concentration, suggesting that the lattice expands due to the substitution of ${\mathrm{Li}}^{+}$ ($r=0.76\AA$, $\mathrm{C}\mathrm{N}=6$) by large radius Tb ions ($r=0.92\AA$, $\mathrm{C}\mathrm{N}=6$). Additionally, the increase of the coordination number of the Ta-O path further certifies the formation of defects in the material.⁵⁰ Furthermore, the local electronic structure of Tb was detected by X-ray absorption near-edge spectroscopy (XANES). Different from the mixed valence state of reference sample Tb₄O₇, the Tb³⁺ activator in ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ is stabilized with the trivalent state [Fig. 4(e) and Fig. S17 in the Supplementary Material], which is similar to the ${\mathrm{Tb}}^{3+}$-doped ${\mathrm{NaLuF}}_4$ nanoscintillators.¹⁶

3.4.

DFT Calculations

To deeply understand how the defect structures affect the NTQ performance of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$, DFT calculations were carried out by constructing ${\mathrm{LiTaO}}_3$ models with these three defects. The structural models are shown in Figs. 5(a1Fig. 2Fig. 3Fig. 4)–5(a3). Differential charge density calculations [Figs. 5(b1)–5(b3) and 5(c1)– 5(c3)] reveal the depletion of electrons at the negative charge defect ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and the accumulation at the positive charge defect ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$. Moreover, there is a significant transfer of electrons from ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ to ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ at the complex defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$. These results manifest that the accumulation and depletion of electrons at the defect sites are significantly increased. And the high electron/positron density in the defects facilitates the ET from the defects to the ${\mathrm{Tb}}^{3+}$ $4\mathrm{f}$ orbital in the luminescence center at high temperatures.^51–53 Therefore, the abundant defects and smooth charge transfer confer the excellent NTQ properties of LiTaO₃:Tb³⁺.

Fig. 5

DFT calculations of ${\mathrm{LiTaO}}_3\text{-}\mathrm{based}$ defect systems. (a1)–(a3) Defect structure models of ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$, ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ in LiTaO₃. (b1)–(b3) (c1)–(c3) Charge density difference of LiTaO₃ with different defects. Yellow represents accumulation of electrons; cyan represents depletion of electrons.

The proposed NTQ mechanism of Tb³⁺-doped LiTaO₃ with triple defects is shown in Fig. S18 in the Supplementary Material. There are three types of defects that can capture carriers (electrons and holes) activated by UV light. The carriers in the defect could be decaptured to compensate for the heat loss upon thermal stimulation, thus achieving negative thermal quenching (NTQ) at high temperatures. Specifically, ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ act as hole traps (h-traps), while ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ acts as an electron trap. Under UV excitation, electrons and holes in the material could be separated, and holes in the VB are captured by the hole trap ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ and $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$, and electrons in the CB are captured by the electron trap ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$, as shown in process (1). In the low-temperature region (300 to 373 K), holes in the shallowest trap ${\mathrm{V}}_{\mathrm{Ta}}^{5-}$ are first decaptured to the luminescence centers (Tb³⁺) [process (2)], and then holes and electrons recombine to generate luminescence [process (3)]. In the mid-temperature region (373 to 425 K), the electrons stored in the trap of ${\mathrm{Tb}}_{\mathrm{Li}}^{2+}$ escape to the excited state level of Tb³⁺ and release energy in the form of luminescence. In the high-temperature region (425 to 475 K), the holes stored in the composite defect $({\mathrm{V}}_{\mathrm{Ta}}{\mathrm{Tb}}_{\mathrm{Li}}{)}^{3-}$ are released to compensate for the heat loss and thus lead to the NTQ behavior.

3.5.

Applications in Temperature-dependent Information Encryption

Since the experimental and computational results indicate that multiple defect levels can trap and detrap optical energy, the ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ phosphor could be very suitable for applications in temperature-dependent information encryption and anti-counterfeiting. To demonstrate information storage capability mediated by triple defects with different trapping depths, we design a temperature-dependent decrypting algorithm, as shown in Fig. 6(a).

Fig. 6

Applications in temperature-dependent decryption, multiple-information storage, and pressure sensing. (a) Schematic diagram of temperature-dependent decrypting process. The information is encrypted by UV irradiation at 298 K. The rise of the temperature to 353 K results in a readout of ${\mathrm{T}}_1$ defect-stored information. After cooling to 298 K, the further increase of temperature to 398 and 453 K successively could decrypt the information stored in ${\mathrm{T}}_2$ and ${\mathrm{T}}_3$ defects, respectively. (b) Schematic diagram of TL curves for temperature-dependent decrypting of triple defects. The dotted line indicates that the encrypted information in the defect is decrypted and read out at the corresponding temperature, and the blue filling represents that the information is still stably stored in the defect. (c) Temperature-dependent decrypting photos of a ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ phosphor wafer with a diameter of 1 cm treated as the process in (a). After UV irradiation for 5 min, the wafer is heated up to 353, 398, and 453 K on a hot plate successively, and treatment at each temperature lasted for 60 s. The information is decoded by luminescence. (d) Schematic diagram of multiple information storage. The ${\mathrm{LiTaO}}_3:1\%{\mathrm{Tb}}^{3+}$ phosphor film is encrypted with the message whale under UV radiation with a mask, and then the charge carriers in the defect are partially emptied by heating at 423 K. The horse pattern is then written at 298 K and is partially emptied at 373 K for 30 s. Finally, encryption of the butterfly pattern was performed at RT. (e) The information pattern decrypted by heating and the corresponding spectrogram. At 298 K, the encrypted message is invisible. The butterfly pattern is decrypted at 353 K, and the horse and whale patterns could be read out at 398 and 453 K, respectively. (f) Schematic diagram of pressure sensing. The cylindrical sample is clipped in the fixture (i), and charged by UV light for 5 min (ii). After UV off and afterglow dissipation (iii), pressure was applied for 0, 5, 10 and 16 s, showing the corresponding pressure sensing images (iv). (g) The pressure sensing curve of intensity versus time obtained by cyclically applying a pressure of 2 kN after UV charging. This indicates that the information stored in the triple defects could be continuously decrypted by applying stress.

First, the optical information storage film consisting of ${\mathrm{LiTaO}}_3:{\mathrm{Tb}}^{3+}$ phosphor and PDMS was encrypted by UV light irradiation at 298 K. Then it was heated to three temperature zones in order to decrypt the information [Fig. 6(a)]. Figure 6(b) shows the TL curves of temperature-dependent decrypting with triple defects. The information decoding photos for the corresponding processes in Figs. 6(a) and 6(b) are shown in Fig. 6(c). The information was decrypted in the form of bright green light emission. At 353 K, the luminescence intensity changed from bright to dimming in 60 s, indicating that trap ${\mathrm{T}}_1$ was emptied first. Upon being continuously heated to 398 and 453 K, the luminous intensity also shifted from bright to dim, manifesting that traps ${\mathrm{T}}_2$ and ${\mathrm{T}}_3$ were also emptied in order. Moreover, after each heating and de-trapping process, the use of UV charging can restore the carriers in the traps. By tailoring the heating duration from 0 to 80 s after UV charging, it can be clearly seen that under thermal stimulation, the stored charge carriers were released in the form of luminescence. The higher the temperature, the faster the charge carriers are released, and the shorter the duration of the thermally excited luminescence (Fig. S19 in the Supplementary Material).

Since the designed pattern information can be encrypted and decrypted intuitively, we also designed a pattern-based encoding method for multiple information encryption [Figs. 6(d) and 6(e) and Figs. S20 and S21 in the Supplementary Material]. First, the whale pattern information was encrypted into the phosphor film using UV light irradiation at 298 K [Fig. 6(d)], followed by a partial readout of the information at 423 K on a hot stage. After cooling to RT, the encryption of the horse pattern was performed and the partial information was emptied again at 373 K. After cooling, the butterfly pattern was encrypted by UV irradiation. The removed information for three patterns is not visible at RT. As the temperature increases from 353, 398, and 453 K, the encrypted patterns butterfly, horse, and whale could be decrypted sequentially [Fig. 6(e)]. The emission spectra at the corresponding temperatures were also recorded. Therefore, high-security and multidimensional information storage using LiTaO₃:Tb³⁺ phosphor with trapping-depth-varied triple defects can be achieved by precisely controlling the heating temperature and encrypting different pattern information.

It is worth noting that the charge carriers stored in the defects also could be stimulated by pressure, which shows potential application of LiTaO₃:Tb³⁺ phosphor in pressure sensing. Figure 6(f) demonstrates that the visual and continuous decryption of information can be achieved by applying pressure to an organic composite block made of LiTaO₃:Tb³⁺ phosphor after 5 min of UV charging. The encrypted information was about 35% of its original intensity after 10 cycles of pressure loading, and the continuous luminescence could be observed by applying pressure onto the composite block [Fig. 6(g)]. Furthermore, the decryption of the information could also be achieved by a 980-nm laser (Fig. S22 in the Supplementary Material). The photo-stimulated luminescence spectrum is completely consistent with the emission spectra recorded under heating and stress, indicating that all this luminescence originates from Tb³⁺. Therefore, the carriers in the defects could achieve a triple thermal, mechanical, and optical response, which provides a new strategy for developing temperature-dependent information encryption, multidimensional information storage, as well as press sensing.