Antithermal Quenching and Multiparametric Temperature Sensing from Mn2+/Tb3+‐Codoped Ca2LaTaO6 Phosphor

Luminescence thermometry plays significant roles in various fields including industrial production, environmental detection, aerospace, and medicine. However, its accuracy improvement remains highly challenging due to the thermal quenching effect of phosphors. Herein, for the first time, a thermal‐activated electron compensation Mn2+‐ and Tb3+‐codoped Ca2LaTaO6 phosphor is developed for multiparametric temperature sensing with tunable emission of Mn2+ and antithermal quenching emission of Tb3+, leading to excellent accuracy at high temperatures. By virtue of the deep electron trap states induced by Mn2+ dopant, the electrons in the deep trap can be thermally activated at high temperatures, which can replenish the attenuated Tb3+ emission caused by thermal quenching, thus bringing out the antithermal quenching phenomenon. On account of the prominent emission properties, the luminescence intensity ratio (LIR) readout and lifetime‐based thermometry are designed, providing a maximum relative sensitivity SR of 3.603% and 1.941% K−1, respectively. Multiparametric temperature sensing and novel data analysis are also employed to further improve the accuracy of the luminescence thermometer. The outstanding relative thermal sensitivity ranging from 8.72% to 16.11% K−1 and temperature uncertainty order of 10−3 are achieved. These results demonstrate that the designed Ca2LaTaO6:Mn2+/Tb3+ phosphor material is a promising thermal‐sensing candidate.

promising luminescent thermometer materials owing to their low toxicity, good physical and chemical stability, and excellent luminescent properties. Different from the extensively reported Mn 4þ emission-based thermometers, Mn 2þ -doped materials are rarely applied in luminescent thermometers, even though they have temperature-sensitive emission. [26] Actually, all the luminescence characteristics (like intensity, bandwidth, spectral position, and lifetime) of Mn 2þ ion are favorable for detecting temperature. [27] This feature caters to the approach of improving the reliability of luminescent thermometers by combining several distinct luminescence parameters with temperature detection capabilities. Therefore, the LIR mode relying on the dual-emission centers of Mn 2þ and Ln 3þ can be optimized to achieve a more reliable multiparameter temperature sensing.
For LIR technique, controlling the antithermal quenching of the emission peak as a reference signal is a feasible and effective way to improve the thermal sensitivity. Previous reports confirm that the antithermal quenching performance in the emission of Mn 4þ , Bi 3þ , and Pr 3þ depends strongly on their electron trap states, which can compensate for the loss of luminescence intensity of the nonradiative relaxed carriers under thermal perturbation. [28][29][30] Moreover, as the phonon energy (or the number of phonons) required for the occurrence of the multiphonon relaxation process during its thermal quenching is larger, Tb 3þ ion is therefore chosen as the second luminescence center, whose emission is more easily complemented by the thermal activated electrons from defect state with elevating temperature.
On account of the above considerations, an innovative Mn 2þ and Tb 3þ -codoped Ca 2 LaTaO 6 (CLTO:Mn 2þ /Tb 3þ ) phosphor with tunable emission of Mn 2þ and antithermal quenching emission of Tb 3þ is for the first time developed. The thermal-activated electron compensation between Mn 2þ and Tb 3þ is evidenced. By virtue of the deep electron trap states caused by Mn 2þ dopant in the bandgap of Ca 2 LaTaO 6 , the electrons in the deep trap can be thermally activated at high temperatures, which can compensate the attenuated Tb 3þ emission resulting from thermal quenching, thus giving rise to the antithermal quenching effect. This particular luminescent properties facilitate the LIR temperature reading which is dependent on the emission of Mn 2þ and Tb 3þ , endowing an extraordinary relative sensitivity S R of 3.603% K À1 . In order to obtain higher sensitivity and reliability, three distinct parameters in the spectra of CLTO:Mn 2þ /Tb 3þ samples are extracted to achieve multiparameter temperature sensing using multiple linear regression (MLR) analysis. Accordingly, the outstanding relative thermal sensitivity ranging from 8.72% to 16.11% K À1 and temperature uncertainty order of 10 À3 are reached. These results confirm the significant potential of CLTO:Mn 2þ /Tb 3þ for luminescence thermometer application. Moreover, the proposed thermal-activated electron compensation mechanism would inspire more deep studies in exploiting highly advanced phosphors for accurate temperature sensing.

Structural and Phase Analysis
The crystal structure of the phosphor has significantly important effects on its optical properties. Figure 1a illustrates the powder X-ray diffraction (XRD) patterns of the as-prepared ion-doped Ca 2 LaTaO 6 (CLTO) samples. Besides, the XRD patterns of the doped CLTO samples with various ion doping concentrations and categories are shown in Figure S1, Supporting Information. All the diffraction peaks of as-prepared samples match well with the standard card (PDF#73-0083) of CLTO, and no impurity phase is observed. The typical perovskite-type Ca 2 LaTaO 6 crystallites belong to the monoclinic phase with the space group P2 1 /n. As depicted in Figure 1b, Ca1 and Ta1 atoms are located at the centers of [Ca1O 6 ] and [Ta1O 6 ] octahedrons, respectively, while La atoms are situated in the interspace of the octahedrons. From Figure 1c, there are two kinds of Ca sites called [Ca1O 6 ] and [Ca2La1O 6 ] octahedrons, which are connected through corners and edges. According to the principle of similar ion radii and equivalent valence, the Mn 2þ and Tb 3þ dopant ions would be inserted in Ca and La sites, respectively. [31,32] Figure 1d-f shows the most intense XRD peak (112) in the 30°-32°range for CLTO:xMn 2þ , CLTO:yTb 3þ , and CLTO:zMn 2þ /5.0%Tb 3þ samples. It is found that the position of the main peaks of all the samples monotonously shifts to higher angles compared with the reference with increasing doping concentration, indicating the lattice contraction. [31] This result is consistent with the above-noted assumption that smaller Mn 2þ (r = 0.83 Å, CN = 6) and Tb 3þ (r = 0.92 Å, CN = 6) ions replace larger Ca 2þ (r = 0.99 Å, CN = 6) and La 3þ (r = 1.05 Å, CN = 6) ions, respectively. To further validate this assumption, the Rietveld refinements on the XRD data of the representative CLTO:1.0%Mn 2þ , CLTO:5.0%Tb 3þ , and CLTO:1.0% Mn 2þ /5.0%Tb 3þ phosphors are conducted to obtain detailed information on crystal structures. Figure 1g-i shows all refinement results with low R-factors (Table S1, Supporting Information), indicating the reliability of the refinement results. The cell volumes of CLTO:1.0%Mn 2þ , CLTO:5.0%Tb 3þ , and CLTO:1.0%Mn 2þ /5.0% Tb 3þ decrease from 272.81 Å 3 (ICSD No. 20 916) to 272. 36,272.47, and 272.16 Å 3 , respectively, implying that the above-noted lattice shrinkage is caused by Tb 3þ and Mn 2þ doping.
X-ray photoelectron spectroscopy (XPS) survey spectrum ( Figure S2a, Supporting Information) illustrates that the characteristic signals of Ca, La, Ta, O, and Mn elements emerge at the corresponding electron binding energy positions. To check the valance state of Mn dopant, the high-resolution XPS spectrum of Mn 2p is performed. The weak peak of Mn in CLTO is due to the low doping amount. The Mn-2p 3/2 peak at 641.13 eV and satellite peak at 646.8 eV coincide with the parameters for representing Mn(2p 3/2 ) spectra for Mn 2þ in Mn oxides. [33] Besides, electron paramagnetic resonance (EPR) spectra of CLTO:Mn 2þ and CLTO:Mn 4þ are shown in Figure S2c, Supporting Information. The resonance signals exhibit the characteristic of an extended exchange-coupled Mn 2þ system. Therefore, both XPS and EPR analyses indicate that the valence state of Mn dopant in CLTO is þ2.
The morphology of the representative CLTO:Mn 2þ /Tb 3þ sample was investigated by scanning electron microscopy (SEM). As shown in Figure S3, Supporting Information, the observed particles exhibiting irregular shapes with sizes of about 50-480 nm (mean size %180 nm) are agglomerated together. The highresolution TEM (HRTEM) images of the CLTO:Mn 2þ , CLTO: Tb 3þ , and CLTO:Mn 2þ /Tb 3þ samples display the interplanar distances of 4.70, 1.69, and 1.61 Å, respectively, indicating the high crystallinity of these samples (Figure 2a). [34,35] The energy-dispersive X-ray spectroscopy (EDX) mapping patterns of the three representative phosphors in Figure 2b demonstrate that the dopant elements are homogeneously distributed in the particles, further elucidating the successful introduction of dopant ions ( Figure S3, Supporting Information). [36]

Photoluminescence Properties
The photoluminescence excitation (PLE) and PL spectra of CLTO:1.0%Mn 2þ sample are shown in Figure S4  www.advancedsciencenews.com www.adpr-journal.com When the Mn 2þ doping concentration over 1.0%, the PL intensities of CLTO:Mn 2þ samples gradually decrease with enhancing Mn 2þ doping amounts due to the concentration quenching effect. Significantly, the emission peaks give a slight redshift with increasing Mn 2þ doping concentration. As shown in Figure 3b, the PL peak position shifts from 680 (x = 0.1%) to 696 nm (x = 10%). According to the Tanabe-Sugano diagram, the lowenergy emission band is usually attributed to Mn 2þ ions undergoing a strong crystal field environment. The crystal field strength (D q ) is mainly dominated by the average bond length (R) between Mn 2þ and the coordination ions, as shown in Equation (1) [32] According to the results obtained in the structural analysis section, the unit cell shrinks with the incorporation of Mn 2þ . Obviously, the crystal field strength increases so that a longer wavelength emission band occurs.
To obtain more information on the concentration quenching of Mn 2þ emission, the room-temperature luminescence decay curves of CLTO:xMn 2þ are measured ( Figure 3c). These decay curves can be well fit by a biexponential equation.  www.advancedsciencenews.com www.adpr-journal.com where I(t) and I 0 denote the emission intensities at times t and zero ms, respectively, A 1 and A 2 are constants, and τ 1 and τ 2 represent the lifetimes of rapid and slow decay, respectively. The average lifetimes (τ) can be calculated by the following formula.
The details are listed in Table S2, Supporting Information. Clearly, the average lifetimes decrease monotonically with increasing Mn 2þ doping concentration because of the gradually increasing probability of nonradiative transitions.
Similarly, the emission of Tb 3þ also exhibits the concentration quenching phenomenon, and the optimal concentration for Tb 3þ doping is 5.0%. The CLTO:yTb 3þ phosphors exhibit a series of intense peaks assigned to the transitions from 5 D 4 to 7 F J ( J = 6, 5, 4, 3) under the excitation of 378 nm (Figure 3d). The optimal doping concentrations of Mn 2þ and Tb 3þ in the codoped CLTO: Mn 2þ /Tb 3þ sample are thus determined to be 1.0% and 5.0%, respectively. The excitation and emission spectra of CLTO:1.0%Mn 2þ /5.0%Tb 3þ sample are shown in Figure 3e,f. For the PLE spectrum monitored at 543 nm, the tail of intense band in the range from 250-300 nm is assigned to the 4f-5d transition and the CTB. [31] The highlighted region in Figure 3e displays several weak peaks centered at 338, 352, 359, 368, and 378 nm, which correspond to the transitions between the 7 F 6 ground state and 5 L 6 , 5 G 4 , 5 D 2 , 5 G 5 , 5 L 10 states. When monitored at 680 nm, a similar excitation spectrum for single-doped CLTO: Mn 2þ phosphor ( Figure S4a, Supporting Information) is presented. When the CLTO:1.0%Mn 2þ /5.0%Tb 3þ sample is under 270 nm excitation (Figure 3f ), the broad red emission band of Mn 2þ can be obviously recognized. However, only the 5 D 4 ! 7 F 6 , 5 D 4 ! 7 F 5 , and 5 D 4 ! 7 F 4 emission peaks for Tb 3þ can be found in the spectrum due to the spectral overlap between the emission broadband of Mn 2þ and the narrow emission for 5 D 4 ! 7 F 3 transition of Tb 3þ . Moreover, the emission intensities for Mn 2þ and Tb 3þ ( 5 D 4 ! 7 F 5 ) are comparable.
It is commonly reported that the increase in thermal stability is associated with the presence of defective states. [4] Thermoluminescence (TL) curves for the as-prepared phosphors are examined and depicted in Figure 4f. As shown, no thermal luminescence glow signal is observed for the CLTO:5.0%Tb 3þ samples. For the CLTO: 0.5%Mn 2þ /5.0%Tb 3þ and CLTO: 1.0%Mn 2þ /5.0%Tb 3þ samples, two clear and partially overlapped peaks are located in the range of 323-523 K ( Figure S7, Supporting Information). It is apparent that the introduction of Mn 2þ into the CLTO:5.0%Tb 3þ lattice produces defect states for trapping carriers. The fitted TL curves with two Gaussian peaks (denoted as P1, P2) are shown in Figure 4g. In detail, the P1 peaks of CLTO:0.5%Mn 2þ /5.0%Tb 3þ and CLTO:1.0%Mn 2þ / 5.0%Tb 3þ are situated at 385 and 382 K, while that of P2 peaks are located at 473 and 439 K, respectively. The trap depth (E T ) could be determined by the following equation. [29,32] where T m is the peak temperature of the TL curve (K). The trap depths are calculated to be 0.77 and 0.76 eV for shallow depth (P1), as well as 0.95 and 0.88 eV for deep trap (P2) of CLTO:0.5%Mn 2þ /5.0%Tb 3þ and CLTO:1.0%Mn 2þ /5.0%Tb 3þ , respectively. As a result, the antithermal quenching of Tb 3þ emission can be explained as follows (Figure 4h): under the excitation of high-energy UV light (e.g. 270 nm), electrons can be easily excited from the ground state to the excited state or even to the conduction band (CB) at room temperature. At the same time, the defect levels with different depths can effectively trap carriers. As the temperature increases, the excited electrons will be released by nonradiative transitions, and the electrons in the shallow trap will be depopulation by thermal activation. As the temperature increases further, the electrons in the deep trap can also be thermally activated. In other words, the electrons in the trap are continuously released to replenish the Tb 3þ emission until the temperature rises to 523 K. This eventually leads to the antithermal quenching phenomenon of Tb 3þ emission, being favorable for the accurate LIR temperature reading, particularly, in high temperatures.

Luminescent Thermometry
Considering the antithermal quenching phenomenon of Tb 3þ emission, CLTO:zMn 2þ /5.0%Tb 3þ (z = 0.1%, 0.5%, 1.0%) is www.advancedsciencenews.com www.adpr-journal.com expected to exhibit great advantages in the LIR temperature sensing. Figure 5a presents the thermal evolution of LIR with temperature from 303 to 543 K, which can be calculated according to the following formula. [29] LIRðTÞ ¼ R 800 640 I Mn ðTÞdλ R 565 535 I Tb ðTÞdλ As a result, the LIR value monotonically decreases with increasing temperature, and it can be deduced as the following equation. [30] LIR where A and B are constants, ΔE is the modified thermal quenching activation energy, and k B is the Boltzmann constant in eV K À1 . As the Mn 2þ concentration decreases from 1.0% to 0.1%, the corresponding fit curves become more linear, and the absolute value of ΔLIR (variation of the LIR in the temperature range of 303-543 K) becomes smaller. Hence, the absolute thermal sensitivity S A can be obtained as the following formula. [5,9]  www.advancedsciencenews.com www.adpr-journal.com The maximum S Amax = 0.056 K À1 (@473 K) in CLTO:1.0% Mn 2þ /5.0%Tb 3þ appears, and all absolute thermal sensitivity curves increase and then decrease with increasing temperature (Figure 5b). The absolute sensitivity is almost equal to a constant over the whole temperature range when the Mn 2þ concentration is 0.1%. To further evaluate the performance of this LIR temperature reading, the thermometric parameter called relative thermal sensitivity (S R ) is calculated as follows. [5,9] The calculated values and corresponding curves are shown in Figure 5c. Likewise, the maximum value of the relative thermal sensitivity S Rmax = 3.603% K À1 (@543 K) is also presented in CLTO:1.0%Mn 2þ /5.0%Tb 3þ . Notably, the values of S R are very close for both CLTO:0.5%Mn 2þ /5.0%Tb 3þ and CLTO:1.0% Mn 2þ /5.0%Tb 3þ at each temperature due to the similar antithermal quenching curves of the Tb 3þ emission in both samples. The optimal CLTO:1.0%Mn 2þ /5.0%Tb 3þ phosphor with outstanding thermal sensitivity demonstrates that such unique antithermal quenching behavior is quite favorable for the LIR temperature reading.
www.advancedsciencenews.com www.adpr-journal.com samples are given in Figure 5d-f and S8, Supporting Information. As presented, the decay curves of all samples decay fast with increasing temperature. Consistently, the emission lifetimes calculated according to Equation (2) and Equation (3) are found to exhibit a monotonic decrease with increasing temperature due to the increased probability of nonradiative transition process. Their lifetimes as a function of temperatures are described by the following equation. [17,19] τ Where τ R (0) is the radiative lifetime at T = 0 K, k B is the Boltzmann constant in cm À1 K À1 , hν is the average energy of phonons coupled to the 4 T 1 ! 6 A 1 transition, 1/τ NR is the nonradiative decay rate, and ΔE Mn is the activation energy of the process. As illustrated in Figure 5g, the fitted curve of CLTO:1.0%Mn 2þ / 5.0%Tb 3þ almost coincides with that of CLTO:0.5%Mn 2þ /5.0% Tb 3þ , being slightly different with that of CLTO:0.1%Mn 2þ / 5.0%Tb 3þ . To assess the performance, the absolute (S a ) and relative (S r ) thermal sensitivities are calculated as follows. [17] S A ¼ dτ dT (10) Figure 5h,i shows that the highest absolute and relative thermal sensitivities are observed in the CLTO:1.0%Mn 2þ /5.0%Tb 3þ sample with maximum values of S Amax = 0.023 K À1 (@470 K) and S Rmax = 1.941% K À1 (@543 K), respectively. Furthermore, it can be recognized in the temperature-dependent absolute and relative thermal sensitivity curves that the relationship between the values near 400 K and Mn 2þ concentration is opposite to that near 500 K. This result may be attributed to the difference in the phonon energy coupled to the electron radiative transition in the host.
Moreover, the temperature uncertainty (δT ) is a thermometric parameter to assess the accuracy of the designed thermometer, which is given by [18,19] δT where δΔ/Δ is the relative error in the determination of the thermometric parameter (δΔ can be determined as a standard deviation in 30 measurements at a certain temperature, as shown in Figure S9, Supporting Information). The temperature uncertainties δT LIR of the LIR temperature readings are higher than 0.067, 0.021, and 0.011 K for CLTO:0.1%Mn 2þ /5.0%Tb 3þ CLTO:0.5% Mn 2þ /5.0%Tb 3þ , and CLTO:1.0%Mn 2þ /5.0%Tb 3þ , respectively. The maximum uncertainty is up to 0.117 K at 303 K for CLTO:0.1%Mn 2þ /5.0%Tb 3þ . The temperature uncertainty δT lifetime of all lifetime-based thermometers is almost below 0.015 K and is as low as 10 À3 orders of magnitude in the temperature range of 423 to 523 K. Although the lifetime-based temperature measurement method has a much lower thermal sensitivity than that of LIR method, it has a significant advantage in terms of temperature uncertainty. In order to obtain more accurate temperature measurement results without sacrificing sensitivity, a novel multiparameter temperature-sensing mode is designed. The relative weights of the full width at half maximum (FWHM), peak energy, and LIR considered by MLR are depicted in Figure 6a. We combine multiple temperature-sensitive parameters in the emission spectrum of Mn 2þ with the earlier-mentioned LIR temperature measurement mode in a reasonable way. Figure 6b presents the temperature-dependent emission spectrum of CLTO:0.5%Mn 2þ /5.0%Tb 3þ in the temperature range of 303-443 K. The peak energies and FWHM of Mn 2þ emission band and LIR parameters serve as independent variables for multiparameter temperature sensing, which can be extracted from this spectrum as a function of temperature, as plotted in Figure 6c-e. Clearly, the dependence of all luminescence parameters on temperature can be described by a linear fitting. These three reliable methods can measure temperature independently and have the maximum relative sensitivities of 0.13, 0.73, and 0.62% K À1 , respectively.
Furthermore, the data of three luminescence parameters were treated by MLR analysis. The temperature could then be expressed as a function of each parameter (Δ 1 , Δ 2 , Δ 3 ) [25] T where β 0 is the intercept, β i (i = 1, 2, 3) is the slope of each thermometric parameter Δ i (Δ 1 , Δ 2 , and Δ 3 represent FWHM, peak energy, and LIR, respectively.), and ϵ is the residual. Hence, the expression for the relative sensitivity S R and the temperature uncertainty δT can be rewritten as follows. [25] S δT As shown in Figure 6f, the calculated temperatures from the MLR agree well with the temperatures measured by the thermocouples, indicating that the method provides highly reliable temperature readings. Notably, compared with the LIR method, a significant improvement in sensitivity over the entire measurement temperature range was observed in the multiparameter temperature sensing through MLR. When the temperature rises from 303 to 443 K, the sensitivity increases from 8.72% to 16.11% K À1 (Figure 6g). More importantly, the temperature uncertainty in the temperature range of 303-443 K reduces to the order of 10 À3 (Figure 6h), which provides a better δT than the lifetime-based thermometer. Such excellent temperature measurement results demonstrate the high advantages and feasibility of CLTO:Mn 2þ /Tb 3þ for multiparameter sensing applications and provide a new approach to develop more reliable and sensitive luminescence thermometers.
Finally, we have compared our results with related phosphors that have been reported in previous literature, and the temperaturesensing performance of our reported Ca 2 LaTaO 6 :Mn 2þ /Tb 3þ phosphor surpasses most of the reported phosphor based thermometers to date (see Table 1), further demonstrating the superior temperature-sensing ability of the obtained antithermal quenching phosphors.

Conclusion
In summary, an innovative Mn 2þ and Tb 3þ -codoped Ca 2 LaTaO 6 phosphor with tunable emission of Mn 2þ and antithermal quenching emission of Tb 3þ is prepared. This unique luminescence phenomenon is attributed to the complementation of thermal-activated electrons at the defect states, which can release electrons by thermal activation to replenish the Tb 3þ emission. The feasibility of the CLTO:Mn 2þ /Tb 3þ phosphors for multiple temperature reading using LIR, emission lifetimes, and distinct luminescence parameters is explored and demonstrated. It is Figure 6. a) Donut chart of the relative weights of the FWHM, peak energy, and LIR considered by MLR. b) Temperature-dependent emission spectra of CLTO:0.5%Mn 2þ /5.0%Tb 3þ phosphors under 270 nm excitation. c-e) Temperature dependence of FWHM, peak energy, and LIR. The lines are the best linear fits of the corresponding data. f ) The ratio of the temperature calculated from the multiparameters sensing through MLR to the temperature measured by the thermocouple. g) Relative thermal sensitivity and h) temperature uncertainty of the multiparameters sensing through MLR.
www.advancedsciencenews.com www.adpr-journal.com concluded that the CLTO:1.0%Mn 2þ /5.0%Tb 3þ provides the maximum relative thermal sensitivity S R of 3.603% K À1 via the LIR approach with the help of Tb 3þ antithermal quenching emission. The lifetime-based luminescent thermometer on CLTO: Mn 2þ /Tb 3þ reveals S R of 1.941% K À1 at 543 K. Multiparametric temperature sensing using MLR is achieved by combining three independently and linearly temperature-dependent luminescence parameters to obtain excellent thermometer performance with relative thermal sensitivity ranging from 8.72% to 16.11% K À1 . Besides, the temperature uncertainty is of the order of 10 À3 . All of these results support that CLTO: Mn 2þ /Tb 3þ is one of the most promising thermal-sensing candidates and provides a new strategy to design and optimize more reliable and sensitive luminescence thermometers.
Characterization: The crystal structures of the prepared samples were characterized by an X-ray powder diffraction (XRD, Rigaku MiniFlex 600 X-ray diffractometer) device employing a Cu target Kα radiation source (λ = 1.54056 Å). The scanning mode for the as-prepared samples was conducted at a diffraction angle range of 10°-80°with a scan rate of 2°min À1 under an operation of 40 kV and 15 mA. Rietveld refinement was performed using the General Structural Analysis System (GSAS) program. The morphology observation and elemental composition analysis of the samples were carried out by SEM (FEI ApreoHiVac) equipped with an X-ray energy-dispersive spectrometer (EDS). Emission spectra (PL), excitation spectra (PLE), and fluorescence decay curves were measured at room temperature using an FS5 fluorescence spectrometer (Edinburgh Instruments) with a 150 W continuous excitation xenon lamp and a pulsed light source. Temperature-dependent emission spectra and decay curves in the range from 303 to 503 K were detected on the same spectrophotometer equipped with a computer-controlled electric heater. XPS was performed on a Thermo ESCALAB 250XI. TL spectrum was performed in a TOSL-3DS TL 3D spectrometer (Rongfan, Guangzhou), using X-rays as the excitation light source.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.