Florescence Intensity Ratio Thermometer in the First Biological Window Based on Non-Thermally Coupled Energy Levels of Tm3+ and Ho3+ Ions

In this study, the up-conversion luminescence and optical temperature sensing properties of Ho3+/Tm3+/Yb3+-co-doped NaLuF4 phosphors were investigated. The visible (475, 540, and 650 nm) and near-infrared light (692 and 800 nm) radiated from 1Ho3+/4Tm3+/Yb3+-co-doped NaLuF4 phosphors were obvious enough for subsequent detection. The slopes in the lnI–lnP plot of the emissions located in the first biological window (650, 692, and 800 nm) were both ∼1.5, which mean that the power had little effect on the three fluorescence peak ratios. Based on the florescence intensity ratios (FIRs) of 650 and 692 nm, the relative sensing sensitivity reaches 0.029 K−1 (476 K). The relative sensing sensitivity based on the FIRs of 800 and 692 nm reaches 0.0076 K−1 (476 K). The results reveal that 1Ho3+/4Tm3+/Yb3+-co-doped NaLuF4 phosphors have potential applications in FIR-based temperature sensing in biological tissue for their high sensing sensitivity. In addition, the emission colors of the sample stabilize in the white light region as the temperature increased from 303 to 467 K, implying that it can also be used in white display.


INTRODUCTION
Temperature (T) is a basic physical parameter in scientific research, production, and living. Therefore, it is crucial to obtain accurate T. Traditional methods of temperature measurement are based on different thermometers, thermal resistance, thermocouples, semiconductor temperature sensor, etc. However, these methods require the physical contact with the object and cannot be used in a corrosive environment or organism. To overcome this issue, numerous types of luminescent thermometers, made up of quantum dots, polymer-based systems, organic dyes, and lanthanide ion (Ln 3+ )-doped materials, have been investigated (Feng et al., 2011;Vlaskin et al., 2010;Yan et al., 2010;Jia et al., 2020). Among these luminescent thermometers, Ln 3+ -doped up-conversion materials have drawn extensive attention in the research of non-invasive temperature measurement due to their fast response, high resolution, low toxicity, remote detection, and so on (Labrador-Páez et al., 2018;Ximendes et al., 2016;Skripka et al., 2017;Nexha et al., 2021). In recent years, the florescence intensity ratio (FIR) from the Ln 3+ -doped up-conversion materials has often been utilized in thermometry, rather than emission lifetime, peak position, and emission color (Runowski et al., 2019;Yuan et al., 2017;Qiu et al., 2020;Du et al., 2018). The FIR temperature sensing method is based on the emissions, radiated from thermally coupled energy levels (200 cm −1 <ΔE<2000 cm −1 ) (Runowski et al., 2019;Yuan et al., 2017;Du et al., 2018) or non-thermally coupled energy levels (ΔE>2000 cm −1 or different Ln 3+ ) (Han et al., 2019;Peng et al., 2021;Chen et al., 2015). Er 3+ ion is promising in FIR temperature sensing for its bright green emission from 2 H 11/2 / 4 S 3/2 and excellent thermal coupling property owing to the suitable energy gap (700-800 cm −1 ), whereas the sensing sensitivity is restricted by the constant energy gap (Xiang et al., 2020;. As far as we know, the sensing sensitivity value is decided by the variation rate of FIR with the temperature. Thus, many strategies are utilized to improve sensitivity, such as choosing the suitable co-doped ions (Lu et al., 2017), changing the suitable host (Huang et al., 2015), distorting local symmetric (Xiang et al., 2020), and using stark sublevels (Suo et al., 2018), while the commonly used fluorescence peaks are located in the visible range, like 520 and 550 nm from Er 3+ , which limit the penetration depth in biological tissues. Therefore, the selected emissions should be located in biological windows to decrease light scattering when they were used in biological tissue temperature sensing (Runowski et al., 2019;del Rosal et al., 2017).
In this study, the emissions (450-850 nm) from Ho 3+ /Tm 3+ / Yb 3+ -co-doped NaLuF 4 phosphors are systemically investigated. High absolute and relative temperature sensitivity can be achieved via choosing suitable FIR. Moreover, the slopes in the lnI-lnP plot of emissions located in the first biological window (650, 692 and 800 nm) are both ∼1.5, which means that power has little effect on the three fluorescence peak ratios. We can take the advantages of these and detect the three emissions for non-invasive temperature measure in biological tissues. Co., Ltd. Ammonium fluoride (AR), sodium hydroxide (AR), oleic acid (AR), and cyclohexane (AR) were purchased from Tianjin Tianli Chemical reagent Co. Ltd. All the chemicals were used directly without further purification.

Sample Preparation
Preparation of NaLuF 4 :1 mol% Ho 3+ , 4 mol% Tm 3+ , and 20 mol% Yb 3+ Phosphors The NaLuF 4 :1 mol% Ho 3+ , 4 mol% Tm 3+ , and 20 mol% Yb 3+ phosphors were prepared using a typical hydrothermal method. The preparation processes of NaLuF 4 : Ho 3+ , Tm 3+ , and Yb 3+ phosphors were as follows. First, calculated amounts of sodium hydroxide were dissolved into 2 ml deionized water. Second, 10 ml absolute ethyl alcohol and 18 ml oleic acid were added in the nitrate solution and then stirred for 5 min at room temperature to form a faint yellow solution. Third, 5 ml aqueous solution was then added, which contained a calculated amount of Lu(NO 3 ) 3 ·6H 2 O, Yb(NO 3 ) 3 ·6H 2 O, Ho(NO 3 ) 3 ·6H 2 O, and Tm(NO 3 ) 3 ·6H 2 O C 6 H 8 O 7 ·H 2 O and 5 ml ammonium fluoride aqueous solution. After stirring for 30 min at room temperature, the mixed solution was transferred into a 50ml autoclave and heated at 180°C for 12 h in a drying oven. After cooling down to room temperature and adding ethanol and cyclohexane, the khaki suspension was centrifuged (8,000 rpm, 2 min) for collection and washed three times with ethanol and deionized water. Finally, phosphors were obtained after drying at 60°C for 10 h.

Instruments
The X-ray diffraction (XRD) pattern of the sample was tested using an X-ray diffractometer (D8-02, Bruker AXS, Germany). The morphology was tested using a transmission electron microscope (TEM: Tecnai G2F 20, FEI, America). The spectra of the samples were tested with an iHR550 grating spectrograph (iHR550, Horiba, France). The 980-nm laser, used to excite the sample, was purchased from Beijing Kipling Photoelectric technology Co., Ltd. The sample was heated by the Linkam THMS 600 heating stage. The temperature of sample was monitored by thermocouple.

XRD Analysis
The XRD pattern of NaLuF 4 : Ho 3+ , Tm 3+ , and Yb 3+ phosphors is presented in Figure 1. We can find that the XRD patterns of the samples can be indexed to a hexagonal NaLuF 4 crystal (the JCPDS standard card no. 27-0726), indicating that the dopants (Ho 3+ , Tm 3+ , and Yb 3+ ions) were successfully incorporated into the host lattice and did not cause significant changes to the crystal structure. The inset showed the TEM images of the sample and the morphology was approximately rod-like. Power-dependent Up-Conversion Luminescence Figure 2 showed the room temperature up-conversion luminescence of NaLuF 4 : Ho 3+ , Tm 3+ , and Yb 3+ phosphors. Under 980-nm laser excitation, the sample exhibited five obvious emission bands centered at 475, 540, 650, 692, and 800 nm. The emissions were attributed to 1 G 4 → 3 H 6 (Tm 3+ ), 5 S 2 / 5 F 4 → 5 I 8 (Ho 3+ ), 1 G 4 → 3 F 4 (Tm 3+ )/ 5 F 5 → 5 I 8 (Ho 3+ ), 3 F 2/ 3 → 3 H 6 (Tm 3+ ), and 3 H 4 → 3 H 6 (Tm 3+ ). It was obvious that the Ho 3+ , Tm 3+ , and Yb 3+ -co-doped NaLuF 4 could provide visible light (475, 540, 650 and 692 nm), which can be used for colorful displaying. Meanwhile, the emissions (650, 692, and 800 nm) located in the first biological window can be utilized to florescent location and temperature detection in biological tissue. The power-dependent up-conversion spectra of the sample are exhibited in Figure 3A, and the emission intensity increased obviously with the power rising. The relationship between laser power (P) and emission intensity (I) can be expressed as I ∝ P n , where n refers to the number of the required photons participated in the up-conversion process (Pollnau et al., 2000). Thus, the value of n can be calculated by the slope obtained from the double logarithmic plot between I and P, as shown in Figure 3B. The values of n were ∼2.1, ∼1.4, ∼1.5, ∼1.5, and ∼1.5, respectively. It implied that the 475, 540, 650, 692, and 800 nm emissions were three-photon, two-photon, two-photon, two-photon, and twophoton processes, respectively. What is more, the value of n for the emissions located in the first biological window was similar to each other, which means that the power has little effect on the three fluorescence peak ratios (Jia et al., 2020). Figure 4 showed the schematic energy levels and the possible up-conversion and energy transfer (ET) processes. Under 980-nm laser excitation, Yb 3+ continuously absorbed 980-nm photons and transfers them to Ho 3+ /Tm 3+ . As described in other research, the possible ET occurred between Tm 3+ and Ho 3+ , inducing the enhancement of 800 nm emission (Wang et al., 2012). Therefore, the FIR of 692/ 800 was fitted through the cubic function rather than the exponential function.

Temperature-dependent Up-Conversion Luminescence Analysis
For FIR-based non-contact thermometry, we studied the temperature-dependent emissions (visible light and the first biological window range) of Ho 3+ , Tm 3+ , and Yb 3+ -co-doped NaLuF 4 powder. The temperature-dependent emission spectra are shown in Figure 5A, and the emission intensities changed as the temperature varied from 303 to 476 K (as shown in Figure 5B). In the visible light range, we choose 475 and 550nm emissions in the FIR-based non-contact thermometry. In addition, the emission colors of the sample stabilized in the white light region as the temperature increased from 303 to 467 K, implying that it can also be used in white display. The detail calculated chromaticity coordinates (x,y) at different temperature are shown in Supplementary Table S1. Figure 6A showed the ratio (R) of blue to green (R 475/550 ) at varied temperature. The temperature dependence of FIR from the non-thermal coupled levels was commonly fitted by polynomial (Lu et al., 2017). Here, we fitted it by mean of a cubic function

Li and Hao
Florescence Intensity Ratio Thermometer FIGURE 4 | Schematic energy levels, possible up-conversion processes, and energy transfer processes in Ho 3+ , Tm 3+ , and Yb 3+ -co-doped materials.

FIGURE 5 | (A)
Up-conversion luminescence of the sample at different temperature, the inset shows the CIE chromaticity coordinates of the sample at different temperature; (B) temperature-dependent integrated intensity of the emissions located at different wavelength. R B 0 + B 1 T + B 2 T 2 + B 2 T 3 , and the fitting parameters are displayed in Supplementary Table S2. To evaluate the sensing capacity of different thermometry method, we calculated the absolute sensing sensitivity (S a ) and relative sensing sensitivity (S r ) through the expression S a dR dT (Jia et al., 2020;Peng et al., 2021). As shown in Figure 6B, the maximum value of S a and S r reached 0.17 K −1 (467 K) and 0.0043 (303 K), respectively. Therefore, the ratio of blue and green can be used for temperature measurement. Figures 7A,B showed the values of R 650/692 and R 800/692 at different temperatures. The values were fitted via cubic function and the fitting parameters are displayed in Supplementary Table  S2. The absolute sensing sensitivity and relative sensing sensitivity are displayed in Figures 8A,B. The maximum values of S a reached 0.042 K −1 (R 650/692 ) and 0.29 K −1 (R 800/692 ) at 303 K, respectively. The maximum values of S r reached 0.028 K −1 at 467 K (R 650/692 ) and 0.0076 at 303 K (R 800/692 ). The ratio of emissions in the first biological window showed higher relatively sensing sensitivity than in the visible range, indicating its extensive application value in temperature measurement, especially in biological temperature measurement. To compare the thermometric capacities of this sample, relevant parameters of other related research studies are listed in Table 1. It can be seen that the sensing sensitivities of our samples are relatively high in these studies. In addition, this method is flexible and practical in thermometry application, such as the selection of detection wavelengths, exciting powers, and detection materials.

CONCLUSION
In conclusion, up-conversion luminescence and temperaturedependent emissions of the transitions 1 G 4 → 3 H 6 (475 nm), 5 S 2 / 5 F 4 → 5 I 8 (550 nm), 1 G 4 → 3 F 4 / 5 F 5 → 5 I 8 (650 nm), 3 F 2/3 → 3 H 6 (692 nm), and 3 H 4 → 3 H 6 (800 nm) from Ho 3+ /Tm 3+ /Yb 3+ -codoped NaLuF 4 phosphors were studied under 980 excitation, and the slopes in the lnI-lnP plot of the emissions were ∼2.1, ∼1.4, ∼1.5, ∼1.5, and ∼1.5, respectively. It implied that the 475, 540, 650, 692, and 800-nm emissions were three-photon, two-photon, two-photon, two-photon, and two-photon processes, respectively.  More importantly, the values of n for the emissions located in the first biological window were similar to each other, which means that power had little effect on the three fluorescence peak ratios. Based on FIRs of 450 and 550 nm, S r reached 0.0043 K −1 . Based on the emissions in the first biological window (R 650/692 and R 800/692 ), S r reached 0.29 and 0.0076 K −1 , respectively. The result reveals that 1Ho 3+ /4Tm 3+ /Yb 3+ -co-doped NaLuF 4 phosphors have potential applications in FIR-based temperature sensing in biological tissue for their high sensing sensitivity. In addition, the emission colors of the sample stabilized in the white light region as the temperature increased from 303 to 467 K, implying that it can also be used in white display.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material; further inquiries can be directed to the corresponding author.