Upconversion luminescence, intrinsic optical bistability, and optical thermometry in Ho3+/Yb3+: BaMoO4 phosphors

Xin Liu; Ruoshan Lei; Feifei Huang; Degang Deng; Huanping Wang; Shilong Zhao; Shiqing Xu

doi:10.3788/COL201917.111601

Abstract

Ho³⁺/Yb³⁺: BaMoO₄ phosphors with different concentrations were fabricated by a gel combustion method. The upconversion (UC) luminescence, intrinsic optical bistability, and the corresponding mechanisms were reported for the present system. The optical thermometric properties based on red (⁵F₅→⁵I₈) and green (⁵F₄/⁵S₂→⁵I₈) emissions were studied. The sensing sensitivities could be tuned by manipulating the cooperative energy transfer process. The highest absolute sensitivity was 99 × 10⁴ K¹ at 573 K, which is larger than that of many previous UC materials.

Rare earth (RE)-doped upconversion (UC) luminescent materials are extensively studied due to their wide applications in bio-imaging, color displays, optical thermometers, optical logic gates, and so on^[1–5]. Particularly, the optical thermometry based on the fluorescence intensity ratio (FIR) of UC materials has attracted growing attentions for its advantages of high spatial resolution, non-contact operation, rapid response, etc., which is suitable for harsh measurement conditions or moving objects^[2,3]. In most cases, the FIR technique determines temperature via two thermally coupled levels (TCLs) of RE ions, which suffers from the disadvantages of low signal discriminability and temperature resolution, owing to the small energy gaps between TCLs^[6,7]. Actually, the FIR between two non-TCLs of RE elements can also be considered as the thermometric parameter. However, the relevant reports are still scarce, not to mention the tunability of sensing behaviors based on non-TCLs.

Besides, intrinsic optical bistability (IOB) is another attractive property observed in some UC materials, in which the emission obeys a hysteresis behavior upon increasing/decreasing pump power. By far, IOB has been investigated in several ${Yb}^{3 +}$ , ${Er}^{3 +} / {Yb}^{3 +}$ , and ${Tm}^{3 +} / {Yb}^{3 +}$ -doped materials, which are suggested to be applicable to optical gates, optical transistors, and so on^[8–10]. However, the IOB process in ${Ho}^{3 +} / {Yb}^{3 +}$ : ${BaMoO}_{4}$ phosphor has not been reported, to our knowledge.

In this Letter, ${BaMoO}_{4}$ is selected as the host material due to its reasonable phonon frequency ( $\sim 800 {cm}^{- 1}$ ), good chemical and physical stability, high anti-humidity characteristics, and so on^[11–13]. The UC luminescence, IOB, and temperature sensing behaviors of ${Ho}^{3 +} / {Yb}^{3 +} : {BaMoO}_{4}$ phosphors were studied. The tunability of sensing sensitivity based on ${}^{5}F_{5}$ and ${}^{5}F_{4} / {}^{5}S_{2}$ levels of ${Ho}^{3 +}$ ions was presented, and its relation with excitation power was discussed.

Sign up for Chinese Optics Letters TOC. Get the latest issue of Chinese Optics Letters delivered right to you！Sign up now

${BaMoO}_{4}$ powders doped with $0.5 mol % {Ho}^{3 +}$ and $x mol % {Yb}^{3 +}$ were prepared by a gel combustion method, where $x = 3$ , 5, 7, and 10. The raw materials of $Ba {({NO}_{3})}_{3}$ (99.99%), ${Na}_{2} {MoO}_{4}$ (A.G.), $Ho {({NO}_{3})}_{3} \cdot 5 H_{2} O (99.99 %)$ , and $Yb {({NO}_{3})}_{3} \cdot 5 H_{2} O$ (99.99%) were dissolved in deionized water, respectively. These solutions were mixed with the appropriate amount, and then glycine was added with the nitrate to glycine ratio of 1:4. The mixed solution was heated at 90°C to form a gel, which was then heated at 400°C for 2 h. Finally, the precursor powders were calcined at 800°C for 2 h to obtain the product. The phase composition of the samples was determined by X-ray diffraction (XRD) analysis (Bruker Axes D2 PHASER Diffractometer) with Cu $K_{α}$ radiation ( $λ = 0.15406 nm$ ). The UC spectra were collected at different temperatures by using a Fluorolog-3/Jobin Yvon spectrofluorometer. A TAP-02 temperature-controlled system was used to modify the sample temperature.

XRD patterns shown in Fig. 1 are evidence that all the diffraction peaks of the samples are well indexed to the tetragonal scheelite-type phase of ${BaMoO}_{4}$ . No extra impurities are detected, suggesting that ${Ho}^{3 +}$ and ${Yb}^{3 +}$ ions are incorporated into the host lattice.

Figure 1.XRD patterns of ${BaMoO}_{4} : 0.5 % {Ho}^{3 +} / x % {Yb}^{3 +}$ ( $x = 3$ , 5, 7, 10).

Figure 2(a) shows the UC luminescence spectra of ${BaMoO}_{4} : {Ho}^{3 +}$ , ${Yb}^{3 +}$ samples under 980 nm excitation. The emission bands peaked at 547, 666, and 755 nm are attributed to ${}^{5}F_{4} / {}^{5}S_{2} \to {}^{5}I_{8}$ , ${}^{5}F_{5} \to {}^{5}I_{8}$ , and ${}^{5}F_{4} / {}^{5}S_{2} \to {}^{5}I_{7}$ transitions, respectively. Meanwhile, the UC emission intensities are enhanced with the increasing ${Yb}^{3 +}$ concentration, owing to the efficient ${Yb}^{3 +} \to {Ho}^{3 +}$ energy transfer (ET) process, and are maximized at $7 mol % {Yb}^{3 +}$ ion concentration. Beyond this, the emissions start to decline due to the concentration quenching effect^[14]. Besides the optimal UC luminescence, an interesting IOB phenomenon is also observed for $0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +} : {BaMoO}_{4}$ . As shown in Fig. 2(b), the green emission intensity does not repeat a linear path with the alteration of excitation power from 300 to 1000 mW and vice versa. The presence of a hysteresis-like curve is evidence for IOB, which is associated with the cooperative ET (CET) process^[8–10,15]. Due to the large concentration of ${Yb}^{3 +}$ ions (7 mol%) compared to that of ${Ho}^{3 +}$ ions (0.5 mol%) in $0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +} : {BaMoO}_{4}$ , large numbers of ${Yb}^{3 +}$ ions in the excited ${}^{2}F_{5 / 2}$ level can form ${Yb}^{3 +} - {Yb}^{3 +}$ dimmers under 980 nm excitation and transfer their energy to ${Ho}^{3 +}$ ions simultaneously via the CET process^[15]. Consequently, an IOB behavior takes place, which may allow the phosphor to be used in optical memory devices, optical gates, and so on.

Figure 2.(a) UC spectra of the samples under 980 nm excitation. (b) IOB plot of green emission for ${BaMoO}_{4} : 0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +}$ phosphor. (c) Logarithmic plots of UC emission intensities as functions of pump power.

In a UC process, the emission intensity $I_{u p}$ is proportional to the pump power ( $P_{N I R}$ ) via the formula $I_{u p} = P_{N I R}^{n}$ , in which $n$ is the number of infrared photons engaged in the UC emission^[16]. From Fig. 2(c), the calculated $n$ values indicate that two photons are involved in the green, red, and deep red UC emission processes. The deviation of the slope values from the expected photon number ( $- 2$ ) may result from the effect of the involved non-radiative channels (such as cross relaxation and multi-phonon relaxation) on the UC processes^[17,18]. Also, as the content ratio of ${Yb}^{3 +}$ to ${Ho}^{3 +}$ is high for $0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +} : {BaMoO}_{4}$ , it is probable for the intermediate levels of ${Ho}^{3 +}$ ions to achieve saturation via absorbing energy of ${Yb}^{3 +}$ ions, which plays a role in the population of the upper emitting levels and leads to reduction of the slope values^[19,20].

Accordingly, the UC mechanisms can be discussed with the help of an energy level diagram (Fig. 3). Since there are no energy levels of ${Ho}^{3 +}$ resonating with 980 nm, the UC emissions of ${Ho}^{3 +}$ ions mainly originate from the ET processes from ${Yb}^{3 +}$ to ${Ho}^{3 +}$ ions via two different ways. The first one is the step-by-step ET processes. Initially, ${Yb}^{3 +}$ ions are excited to the ${}^{2}F_{5 / 2}$ level via the ground state absorption (GSA) process. Then, the ${Ho}^{3 +}$ ions in the ${}^{5}I_{8}$ level are pumped to ${}^{5}F_{4} / {}^{5}S_{2}$ levels through the ET-1 and ET-2 processes. After that, the radiative transitions from the ${}^{5}F_{4} / {}^{5}S_{2}$ levels to the ${}^{5}I_{8}$ and ${}^{5}I_{7}$ ones lead to 547 and 755 nm emission, respectively. Alternatively, some ${Ho}^{3 +}$ ions at the ${}^{5}I_{6}$ level may also relax non-radiatively to the ${}^{5}I_{7}$ level and are further elevated to the ${}^{5}F_{5}$ level via the ET-3 process. Finally, the 666 nm emission occurs via the ${}^{5}F_{5} \to {}^{5}I_{8}$ transition. In the second way, the CET process from the ${Yb}^{3 +}$ to ${Ho}^{3 +}$ ions contributes to the populations of the ${}^{5}F_{2} / {}^{3}K_{8}$ levels. Then, these ions non-radiatively relax to the ${}^{5}F_{4} / {}^{5}S_{2}$ levels and finally generate the 547 and 755 nm emissions.

Figure 3.Energy level diagram and the possible UC processes under 980 nm excitation.

To investigate the optical thermometric behaviors, Fig. 4(a) illustrates the temperature-dependent UC emission spectra of ${BaMoO}_{4} : 0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +}$ at 100 mW, which are normalized to the 666 nm peak. With temperature increment, the 547 nm emission intensity decreases gradually compared to the 666 nm emission. One of the reasons is the promoted multi-phonon relaxation from the ${}^{5}F_{4} / {}^{5}S_{2}$ to the ${}^{5}F_{5}$ level at elevated temperatures^[21,22]. Additionally, the increment of temperature also accelerates the ${}^{5}I_{6} \to {}^{5}I_{7}$ non-radiative relaxation process, which is in favor of the 666 nm emission rather than the 547 nm one (Fig. 3). Hence, the 547 and 666 nm emissions are potentially for FIR thermometry due to their diverse thermal response, which has an excellent signal discriminability ( $- 3266 {cm}^{- 1}$ ). The variations of FIRs ( $I_{666} / I_{547}$ ) in the 298–573 K temperature range dependent upon different excitation powers are presented in Fig. 4(b). Clearly, the FIRs keep growing with rising temperatures, and the relation between FIR and temperature can be well fitted by the following equation: $F I R = \frac{I_{666}}{I_{547}} = A \exp (\frac{- B}{T}) + C,$ (1)where $A$ , $B$ , and $C$ are the constants related to the research system. As shown, perfect fittings can be achieved with the correlation coefficient $R$ of up to −0.998. Therefore, the temperature can be detected based on the intensity ratio between the ${}^{5}F_{5} \to {}^{5}I_{8}$ and ${}^{5}F_{4} / {}^{5}S_{2} \to {}^{5}I_{8}$ transitions.

Figure 4.(a) Temperature-dependent UC spectra. (b) The FIRs as functions of temperature under the different excitation powers. (c) The variations of $S_{a}$ values with temperature and excitation power.

Then, the key parameter of absolute sensing sensitivity ( $S_{a}$ ) can be evaluated, which is defined as $S_{a} = | dFIR / d T |$ ^[23]. Figure 4(c) shows the maximal $S_{a}$ value decreases from $99 \times 10^{- 4} K^{- 1}$ to $48 \times 10^{- 4} K^{- 1}$ with an increase of excitation power from 100 to 750 mW. This may be explained as follows: a high thermal sensitivity based on the ${}^{5}F_{5}$ and ${}^{5}F_{4} / {}^{5}S_{2}$ levels requires the efficient population of the ${}^{5}F_{5}$ level and depopulation of the ${}^{5}F_{4} / {}^{5}S_{2}$ levels with temperature. However, the CET process becomes more and more effective with increasing excitation power^[24], which is in favor of the population of ${}^{5}F_{4} / {}^{5}S_{2}$ levels rather than the ${}^{5}F_{5}$ level. Consequently, FIRs ( $I_{666} / I_{547}$ ) decline with the increasing excitation power [Fig. 4(b)], indicating a reducing population rate ratio between the ${}^{5}F_{5}$ and ${}^{5}F_{4} / {}^{5}S_{2}$ levels. Hence, the pump power increment is harmful for the thermal response of FIRs, resulting in decreasing sensitivity. The results reveal that the excitation power affects the thermometric performance remarkably, which has normally been ignored in past investigations.

In summary, a series of ${Ho}^{3 +} / {Yb}^{3 +} : {BaMoO}_{4}$ phosphors have been prepared and studied. Upon 980 nm excitation, ${BaMoO}_{4} : 0.5 % {Ho}^{3 +} / 7 % {Yb}^{3 +}$ phosphor exhibits the most intense UC emissions and an IOB behavior. Based on FIR between the ${}^{5}F_{5}$ and ${}^{5}F_{4} / {}^{5}S_{2}$ levels, the thermal sensitivities are found to depend on the change of excitation powers due to the impact of the CET process. Moreover, the phosphor has relatively high thermal sensitivity ( $99 \times 10^{- 4} K^{- 1}$ ) and excellent signal discriminability. The results suggest that the prepared phosphor may be applied in an optical thermometer, optical bistable device, and some other UC-based devices.