1. Introduction

Compared with the traditional contact thermometers, optical temperature sensing owns many prominent advantages, such as the quick response, high spatial resolution and its non-contact detection mode [1,2]. These features make temperature sensing promising for many complex and special cases in which the traditional thermometers are powerless. In particular, by virtue of the quick response, temperature sensing of fast-moving objects and identification of rapid variation in temperature can be realized. High spatial resolution in submicron scale can be achieved with the help of microscopy using luminescent nanoparticle as temperature probe [2]. In recent several years, much attention has been paid to temperature sensing for biological tissue or cells using the temperature related upconversion luminescence due to its unique advantages in better penetration for biological tissue of the 980 nm excitation and minimized background autofluorescence [3]. However, upconversion luminescence suffers from low quantum yields and the laser heating effect caused by the strong absorption of Yb³⁺ ions, which is harmful to temperature sensing applications. Therefore, in many situations except for biological applications, adopting downshifting luminescence with high quantum yields and less light heating effect to detect temperature is promising alternative to upconversion.

The relative temperature sensitivity S_R is a significant parameter to show the performance of the temperature sensors. There have been great concerns on the improvement of S_R by exploring novel sensing materials and novel sensing mechanisms in order to developing high-performance temperature sensors. During the past several years, the sensing technique based on the temperature dependent fluorescence intensity ratio (FIR) of two thermally coupled energy levels (TCELs) has been widely investigated and has been regarded as a promising technique for optical temperature sensing application [3–5]. The S_R value of such method is in proportion to the energy gap of the TCELs. However, the energy gap was limited in order to achieve thermal coupling because larger energy gap will lead decoupling effect, which consequently restricts the S_R value [5–7]. This intrinsic drawback makes further promoting S_R impossible in theory. To the best of our knowledge, the highest S_R value based on the FIR technique using TCELs is 2805/T² using ⁴F_7/2 and ⁴F_3/2 of Nd³⁺ [8]. Therefore, great efforts have been made to search for high sensitive optical thermometry [7,9–20]. For instance, novel strategies based on the dual emitting centers (such as Eu³⁺/Nd³⁺ [7], Eu³⁺/Tb³⁺ [9], Eu³⁺/Eu²⁺ [10], Eu³⁺/ligand [11], Eu³⁺/Mn²⁺ [12], Eu²⁺/Cr³⁺ [13]), the metal-to-metal intervalence charge transfer state [14,15], the thermal population of lower-lying excited energy levels (such as Eu³⁺ [16], Sm³⁺ [17]) have been evidenced to be applicable for optical temperature sensing. Even so, novel materials or strategies with higher S_R are still needed to be explored for potential application.

Recently, we reported the temperature induced shift of the charge transfer band (CTB) edge in rare earth ions doped vanadates [21]. In detail, with increase of temperature, the CTB edge shifts to longer wavelength remarkably, leading to enhanced excitation efficiency at a certain wavelength in the tail of the CTB. That is to say, under such excitation, the fluorescence intensity will increase rapidly with temperature increasing. Based on this point, if rare earth ions possessing TCELs are introduced, we expect that the fluorescence of the upper-level in the TCELs will increase more dramatically with temperature due to the Boltzmann distribution in premise of no temperature quenching, which may be used for intensity-based optical thermometry with high temperature resolution. Herein, Sm³⁺ and Er³⁺ doped LuVO₄ powder samples were investigated to evaluate the feasibility of such strategy. The temperature dependent fluorescence from the TCELs of Sm³⁺ and Er³⁺ were investigated under a specific excitation which was selected according to the temperature dependent CTB. The obtained temperature dependences indicated their potential application in optical temperature sensing.

2. Experiments

The LuVO₄:4% Sm³⁺ and LuVO₄:2% Er³⁺ powder samples were prepared by high temperature solid state method [21]. The raw materials including Lu₂O₃, Sm₂O₃ and Er₂O₃ are all of 99.99% purity, and V₂O₅ is of analytical grade. X-ray diffraction (XRD) analysis was performed using a powder diffractometer (XʹPert³). The optical measurements of the as-prepared samples including excitation spectra and emission spectra as well as fluorescence decay curves at different temperatures were carried out using a spectrofluorometer (HORIBA Florolog-3) equipped with a 450 W xenon lamp and a pulsed 370 nm spectraLED (HORIBA). The temperature variation of the powder was controlled using a temperature control device using a heating tube and a temperature controller (OMRON E5CC-800) equipped with a type-K thermocouple. The powder sample was pressed tightly into a circular groove in a temperature controlled copper plate. It is noted that the amount of the powder sample was little and the sample was attached as a very thin layer to guarantee that the sample temperature is the same as for the sensor.

3. Results and discussion

The crystal structures of the as-prepared powder samples were identified by XRD. Figure 1 gives the obtained XRD patterns of the samples, which agree well with the standard tetragonal LuVO₄ (JCPDS No. 82-1977) diffraction data, indicating that the samples we synthesized are highly crystalline pure tetragonal LuVO₄ phase.

 

Fig. 1 XRD patterns of the as-prepared LuVO₄:4% Sm³⁺ and LuVO₄:2% Er³⁺ powder samples and the standard tetragonal LuVO₄ diffraction data.

Download Full Size | PDF

First, we investigated the temperature dependent fluorescence properties of LuVO₄:4% Sm³⁺ according to the strategy stated above to evaluate the possibility for its potential application in optical thermometry. The excitation spectra of LuVO₄:4% Sm³⁺ were recorded at different temperatures from 300 to 480 K by monitoring 602 nm emission, which were shown in Fig. 2. The broad excitation band is attributed to the V-O charge transfer process and the followed efficient energy transfer to Sm³⁺ ions. It is clearly that the CTB edge shifts to longer wavelength with temperature increases, which was mainly caused by the thermal population of the vibrational sublevels of the ground electronic energy level of VO₄^3- group [22]. As a result of the temperature induced red shift, the excitation intensity in the tail of the CTB enhances with temperature dramatically, especially around 360 nm. Therefore, 360 nm was selected as the excitation wavelength for the measurements of emission spectra at different temperatures from 300 to 480 K as shown in Fig. 3. It is noted that the spectra ranging from 520 to 550 nm were recorded using a larger slit width compared with that ranging from 550 to 700 nm due to its relatively weak emission intensity. The emission peak at 537 nm is ascribed to ⁴F_3/2 → ⁶H_5/2 transition of Sm³⁺, while the emission peaks at 565 nm, 602 nm and 647 nm are ascribed to ⁴G_5/2 → ⁶H_5/2, ⁶H_7/2 and ⁶H_9/2 transitions, respectively. As we expected, all the emission intensities increase dramatically with the increase of temperature. And the intensity variation from ⁴F_3/2 is more dramatic than that from ⁴G_5/2, which is consistent with the Boltzmann distribution of ⁴F_3/2 and ⁴G_5/2 [23,24].

 

Fig. 2 Excitation spectra (λ_em = 602 nm) of LuVO₄:4% Sm³⁺ recorded at different temperatures from 300 to 480 K.

Download Full Size | PDF

 

Fig. 3 Emission spectra (λ_ex = 360 nm) of LuVO₄:4% Sm³⁺ recorded at different temperatures from 300 to 480 K.

Download Full Size | PDF

Figure 4 shows the temperature dependences of the intensity I_520-545 integrated from 520 to 545 nm and the intensity I_550-679 integrated from 550 to 679 nm, which represent the emissions originating from ⁴F_3/2 and ⁴G_5/2, respectively. The relationships between the integrated intensities and temperature were well fitted using Arrhenius-type equation as shown in Fig. 4. In order to reduce the experimental error, the data of I_520-545 at 300 K was not used because it is very weak, which can be seen from Fig. 3.

 

Fig. 4 Temperature dependences of the integrated intensities I_520-545 and I_550-679 originating from ⁴F_3/2 and ⁴G_5/2 of Sm³⁺, respectively.

Download Full Size | PDF

The excitation spectra of LuVO₄:2% Er³⁺ were recorded at the same temperature range by monitoring the 554 nm emission originating from ⁴S_3/2 of Er³⁺. As shown in Fig. 5(a), the CTB edge exhibits the same temperature dependent shift as in Fig. 2. Meanwhile, the integrated intensities of the CTB and the ⁴I_15/2 → ⁴G_11/2 excitation transition of Er³⁺ decrease remarkably. This is mainly caused by the decreased population of the ⁴S_3/2 with the increase of temperature due to the Boltzmann distribution of ²H_11/2 and ⁴S_3/2 of Er³⁺. Under the excitation of 360 nm, the emission spectra of LuVO₄:2% Er³⁺ were recorded at different temperatures from 300 to 480 K, which are shown in Fig. 5(b). With the increase of temperature, the characteristic green emissions of Er³⁺ originating from the TCELs ²H_11/2 and ⁴S_3/2 increase remarkably. Figure 5(c) shows the temperature dependences of the integrated intensities I_515-540 and I_540-565 corresponding to the ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 transitions, respectively. As expected, the emission originating from ²H_11/2 as the upper-level in the TCELs exhibits more dramatic temperature dependence. The experimental data were also well fitted with Arrhenius-type equation as shown in Fig. 5(c). Moreover, the temperature dependence of the ratio of I_515-540 to I_540-565 is also given in Fig. 5(d), which has been widely investigated for FIR-based optical thermometry previously and was used for comparison with our proposed strategy here. The energy gap between ²H_11/2 and ⁴S_3/2 was obtained to be 751 cm⁻¹ according to the fitting result.

 

Fig. 5 (a) Excitation spectra (λ_em = 554 nm) and (b) emission spectra (λ_ex = 360 nm) of LuVO₄:2% Er³⁺ recorded at different temperatures from 300 to 480 K. (c) Temperature dependences of the integrated intensities I_515-540 and I_540-565 originating from ²H_11/2 and ⁴S_3/2 of Er³⁺, respectively. (d) Temperature dependent ratio of I_515-540 to I_540-565 for LuVO₄:2% Er³⁺.

Download Full Size | PDF

With the increase of temperature, another crucial factor influencing the emission intensity variation is temperature quenching, which cannot be ignored. Therefore, the temperature dependent decay curves of the as-prepared samples were measured by monitoring their respective emissions at the same temperature range. As shown in Fig. 6, the fluorescence lifetime of Er³⁺ decreases with the increase of temperature, while no obvious variation exists for that of Sm³⁺, indicating that temperature quenching weakens the emission of Er³⁺ in our investigated temperature range. We conclude that these different temperature quenching behaviors are relevant to the intermediate energy levels of Er³⁺ and Sm³⁺. From the energy level diagrams of Er³⁺ and Sm³⁺, the energy gap between the luminescent energy levels of Er³⁺ (²H_11/2 and ⁴S_3/2) and the adjacent lower energy level ⁴F_9/2 is obviously smaller than that between the luminescent energy levels of Sm³⁺ (⁴F_3/2 and ⁴G_5/2) and the adjacent lower energy level ⁶F_11/2. Therefore, temperature quenching of Er³⁺ emission through multi-phonon relaxation process occurs more easily than that of Sm³⁺ emission.

 

Fig. 6 Temperature dependent fluorescence decay curves for (a) LuVO₄:2% Er³⁺ (λ_em = 526 nm) and (b) LuVO₄:4% Sm³⁺ (λ_em = 602 nm).

Download Full Size | PDF

The relative sensitivities S_R, defined as the relative change of the detection signal with respect to temperature variation, of the above investigated sensing strategies are shown in Fig. 7. Obviously, the S_R values based on our proposed strategy are remarkably higher than that based on the FIR method using the TCELs of Er³⁺. And the emission from the upper-level in the TCELs exhibits higher S_R than that from the lower-level. By virtue of the temperature dependent emission originating from ⁴F_3/2 as the upper-level in the TCELs of Sm³⁺, we obtained a very high S_R of 4304/T² in the temperature range of 330-480 K. And the maximum reaches up to 3.95% K⁻¹ at 330 K. As we mentioned in the introduction, the highest S_R value of the FIR technique using TCELs was reported to be 2805/T² [8]. Obviously, our proposed strategy is superior to that method, indicating a promising approach for high performance optical thermometry.

 

Fig. 7 The obtained relative sensitivity S_R in the investigated temperature range.

Download Full Size | PDF

4. Conclusion

In summary, by virtue of the enhanced excitation efficiency in the tail of the CTB of VO₄^3- group caused by the temperature induced red shift of the CTB edge, the emissions from the TCELs of Sm³⁺ (⁴F_3/2 and ⁴G_5/2) and Er³⁺ (²H_11/2 and ⁴S_3/2) in LuVO₄ increase dramatically with the increase of temperature from 300 to 480 K under the excitation of 360 nm. And the emission from the upper-level in the TCELs exhibits more dramatic temperature dependence than that from the lower-level due to the Boltzmann distribution of the paired levels. As a result, based on the temperature dependent emission of the upper-level in the TCELs of Sm³⁺, high relative sensitivity was obtained to be 4304/T². The maximal S_R value reaches 3.95% K⁻¹ at 330 K, which is remarkably superior to that based on the FIR method using TCELs. Such excellent performance makes the as-proposed strategy a promising candidate for optical thermometry with high temperature resolution.