1. Introduction

Optical temperature sensors based on upconversion (UC) luminescent materials with rare earth (RE) ions as the activators have attracted much interest in recent years [1]. Generally, each of the following parameters: the fluorescence intensity, the luminescence lifetime, and the fluorescence intensity ratio (FIR) of luminescent materials can be used for temperature measurement by monitoring their change induced by temperature. Among them, the FIR technique [1–6] was regard as one of the promising techniques because it can reduce the dependence of the measurement conditions, such as fluorescence loss, fluctuations of exciting light and electromagnetic compatibility problem. FIR technique requires the luminescent material to display two emission bands whose response to temperature has to be different. Based on the special 4f energy levels, rare earth activators have been extensively adopted as temperature measurement when they are doped into appropriate hosts. There are two modes of rare earth levels three-level system and four-level system can be used for optical temperature sensor based on the UC. Two thermal coupling levels (2, 3) and a ground state level (1) are involved in the three-level system (E₃>E₂>E₁). The two thermal coupled levels transited to the ground state energy level and emit two green lights. The intensity ratio (I₃₁/I₂₁) of the two green light changes regularly with temperature. This model is suitable for most optical temperature sensing materials. The four-level system, was introduced by Haro-González et al [7], is mainly used for Ho ions doped optical temperature sensing materials. It contains two upper levels (E₃, E₄) and two lower levels (E₁, E₂). The energy of four levels arrange in the order of E₄>E₃>E₂>E₁. The FIR can be concluded by the formula of $\frac{{\text{I}}_{42}+{\text{I}}_{32}}{{\text{I}}_{41}+{\text{I}}_{31}}$ . Among all the rare earths, Er³⁺ ion is one of preferable choice as temperature sensor material [5,8–14]. The two green emission bands originated from ²H_11/2 → ⁴I_15/2 and ⁴S_3/2→ ⁴I_15/2 transitions of Er³⁺ could be used for measuring temperature which be analyzed in three-level system. Moreover, due to its large absorption cross section at 980 nm and efficient energy transfer from Yb³⁺ to Er³⁺, Yb³⁺ ions are usually used as the co-doped ion to sensitize Er³⁺.

In order to obtain a good optical temperature sensor, the factor of host matrix has to be taken into account as well. It is known that the optical temperature sensors mainly focus on the host materials based on glasses and fluorides [5, 11, 15], from which the upconversion luminescence originated at early times. However, the relatively poor stability and toxicity greatly hamper their further application for temperature sensor [16]. Compared with glasses and fluorides, oxides can overcome these drawbacks due to their excellent thermal and chemical stability and environmental benefits [16]. Taking into account these factors and the research foundation of our research group in the field of oxide ferroelectric, most of our previous studies have focused on rare earth doped high temperature ferroelectric materials. Although it is suitable for high temperature applications, the value of optical temperature sensitivity is not as high as expected which may be due to the high phonon energy of the ferroelectric material. As we known, ZnWO₄ has a wolframite structure which belongs to a low crystal symmetrical monoclinic crystal system with the space group P2/c (13) [17–19]. Meanwhile, ZnWO₄ has some other advantages that benefit UC, such as a high chemical and physical stability and low phonon threshold energy (199.5 cm⁻¹) [20,21]. All of these above make ZnWO₄ as one of the potential luminescent materials. When selected it as a matrix, RE ions (Er, Dy, Ho, Eu) doped ZnWO₄ materials showed very good fluorescence properties [22–28]. However, no studies are available on the UC luminescence, optical temperature sensing properties and the solubility for RE ions doped ZnWO₄ materials. Taking into account these advantages of ZnWO₄, we expect that ZnWO₄ could act as an ideal host for UC luminescence and optical temperature sensor.

In this work, we prepared Er³⁺/Yb³⁺ co-doped ZnWO₄ phosphors through a facile solid-state reaction method. The materials exhibited strong visible emission through UC (green and red emission under near infrared excitation) processes. Meanwhile, color-tunable UC emissions from green to red were realized due to the variety of the red-to-green ratio caused by changing the doping concentration of Er³⁺/Yb³⁺, which endows this kind of material with potential applications in multiple biolabels. The optical temperature sensing properties of the samples were studied in the range of 83-583 K by means of FIR techniques. The results demonstrate that this material is a promising candidate for high sensitive temperature sensor.

2. Experimental

The Er³⁺/Yb³⁺ co-doped ZnWO₄ phosphors (ZnWO₄: Er³⁺/Yb³⁺) were prepared by a solid state reaction method. High purity (>99.9%) raw materials ZnO and WO₃ were mixed at 1:1 molar ratio, and x mol% Er₂O₃ and y mol% Yb₂O₃ were added (x = 0.025, 0.05, 0.1, 0.2, 0.5, 1.0, 1.3; x:y = 1:10) [13,14]. The initial powders were ground finely in an agate mortar, then dried and calcined at 750 °C for 5 h. After calcination, the ground powders, added with 10 wt. % polyvinyl alcohol (PVA) binder, were pressed into disk shaped pellets of 10 mm in diameter and about 2 mm in thickness. Finally, samples were sintered at temperatures of 1120 °C for 4 h in an alumina crucible in air.

The crystal structure was identified by an X-ray diffractometer (XRD) (D/MAX 2550, Rigaku, Japan) with Cu Kα1 radiation (λ = 0.154056 nm), tube voltage 40 kV, tube current 40 mA. The XRD profiles of the samples were collected in the range of 10° ≤ 2θ≤ 90°, with a step size of 0.02° and a step time of 3 s. The luminescence spectrum collected from the samples was recorded by using a fluorescence spectro-fluorometer (F-7000, Hitachi, Japan). For UC measurement, a 980 nm laser diode (LD) (HJZ980-100) with keeping the power at 30 mW was used to excite the surface of ceramic samples. The sample temperature was measured by a Pt-100 thermocouple located at a heating stage controlled by a TP94 temperature controller (Linkam Scientific Instruments Ltd, Surrey, UK). The energy dispersive X-ray spectrum (EDS) was examined using field emission scanning electron microscopy (FEI Nova Nano SEM 450, America).

3. Results and discussion

3.1 Color tuning of UC luminescence

Under 980 nm near infrared laser excitation, the corresponding UC emission spectra which were measured under similar conditions and digital photographs of the Er³⁺/Yb³⁺ co-doped ZnWO₄ samples are shown in Fig. 1(a). Similar to other Er³⁺/Yb³⁺ co-doped photoluminescence materials [29–31], it can be clearly seen that the typical UC emission consists of three parts: red emission band at about 640~680 nm, attributed to the Er³⁺ ions’ ⁴F_9/2→⁴I_15/2 transition; two green emission bands at 549 and 524 nm, attributed to the Er³⁺ ions’ ⁴S_3/2→⁴I_15/2 and ²H_11/2→⁴I_15/2 transitions, respectively. The UC mechanisms are illustrated in Fig. 2 (GSA denotes ground state absorption; ESA denotes excited state absorption). Under a 980 nm laser excitation, the Er³⁺ ions are firstly excited from the ⁴I_15/2 to ⁴I_11/2 by GSA. And then by energy transfer process from the Yb³⁺ ions (it is dominant) and GSA/ESA processes, the ⁴F_7/2 states of Er³⁺ are populated. Next, the ²H_11/2 and ⁴S_3/2 levels are populated by a nonradiative relaxation from the ⁴F_7/2 level. Finally, green emissions between 524 nm and 549 nm are produced by the ²H_11/2 →⁴I_15/2 and ⁴S_3/2 →⁴I_15/2 transitions, respectively. However, there are two possible UC channels responsible for the red emission. Firstly, the nonradiative relaxation from ⁴S_3/2 to ⁴F_9/2 can contribute to the red emission [shown as 1 in Fig. 2]. This channel requires a lower concentration of Er³⁺. Secondly, parts of Er³⁺ at ⁴I_11/2 level relax to the ⁴I_13/2 level. Then the long-lived ⁴I_13/2 level is promoted to ⁴F_9/2 level through the ESA or energy transfer [shown as 2 in Fig. 2]. The Yb³⁺ ions enriched the population of all the emitting levels via a suitable energy transfer process that amplify the emissions correspondingly.

 

Fig. 1 (a) The UC luminescence spectra and corresponding luminescence photographs of ZnWO₄: Er³⁺/Yb³⁺ samples, (b) Red-to-green emission intensity ratios as a function of doping concentration of Er³⁺.

Download Full Size | PDF

 

Fig. 2 Schematic diagram of Er³⁺ energy levels for visible UC photoluminescence mechanisms in ZnWO₄: Er³⁺/Yb³⁺ phosphors.

Download Full Size | PDF

It can also be seen that the locations of the emission peaks do not shift obviously, while the relative intensity of green and red emission changed with increasing doping concentration of Er³⁺/Yb³⁺. When a small amount of Er³⁺/Yb³⁺ (0.025≤ x ≤0.05) was introduced into the ZnWO₄ crystal lattice, the red-to-green emission intensity ratio decreased, resulting in a bright green UC emission (x = 0.05 mol%) which can be used for optical temperature sensors and will be discussed later. When we further increased the amount of Er³⁺/Yb³⁺, the red-to-green ratio rapidly increased, as shown in Fig. 1(b). The peaks in the red emission region are much stronger than that in the green emission region, so a red color was seen with naked eyes [photographs in Fig. 1(a)]. From these results, we can conclude the optimal doping concentration is at x = 0.05 mol% for the green emission, whereas for the red one it is 1.0 mol%. It is known that when the doping concentration reaches a certain degree, the distance between Er³⁺ ions became small and a fraction of energy migrate to the quenchers. This variation in color of the samples has a great relationship with the high concentration of Er³⁺/Yb³⁺. At high Er³⁺/Yb³⁺ concentrations, the cross-relaxation between ²H_11/2 and ⁴I_15/2 (²H_11/2 + ⁴I_15/2 → ⁴I_9/2 + ⁴I_13/2) [shown as CR1 in Fig. 2] is enhanced, which contributes to the concentration quenching for green emission. Meanwhile, this process leads to population of the ⁴I_13/2 level from which the ⁴F_9/2 level can be populated by the ESA or energy transfer processes. Furthermore,the energy transfers from Yb³⁺ to Er³⁺ are highly efficient, which results in great population of Er^{3+ 4}I_11/2 and ⁴F_7/2 states. The cross-relaxation of ⁴F_7/2 + ⁴I_11/2-⁴F_9/2 + ⁴F_9/2 [shown as CR2 in Fig. 2] between Er³⁺ ions becomes significant and dominant for realizing the population of the Er³⁺: ⁴F_9/2 excited state from which red emission is yielded. This is why the red color dominant emission can be obtained when the Er³⁺ concentration surpasses 0.05 mol%. Moreover, the process of ⁴I_13/2→⁴F_9/2 is very efficient and result in bypassing the green ⁴S_3/2 state. So, the intensity of red emission increases. The emission intensity of red light reaches their maximum at x = 1.0 mol%, then decreases contrarily with increasing Er³⁺ content, owing to the concentration-quenching effect. The change trend of green to red light intensity is also observed by Yang et al [32].

Figure 3(a) displays the UC emission spectra of the ZnWO₄: 0.05 mol% Er³⁺/0.5 mol% Yb³⁺ phosphor upon a 980 nm excitation at different pump powers in the range of 40 – 240 mW. It can be found that there were no significant changes in the UC spectra, while the UC emission intensity showed an upward trend with the increment of pump power. The LnI – LnP plot, between the input (excitation) power and emission intensities of the different bands, tells us about the number of incident photons involved in a particular UC transition. This also helps us to understand the excitation mechanisms involved in the UC process. For the ZnWO₄: Er³⁺/Yb³⁺ phosphor with x = 0.05 mol % [see the inset of Fig. 3(a)], the value of the slope for the green emission bands at 524 nm, 549 nm were observed to be 1.81 and 1.87, and the red emission band at 668 nm were observed to be 1.70, respectively. The slope values indicate that two photons are involved in both green and red emission processes. Furthermore, the colour coordinates were calculated at different pump power as shown in the Commission Internationale de l’Eclairage (CIE) chromaticity diagram [Fig. 3(b)]. The colourcoordinates were found to be almost constant at the pure green region when increasing the excitation power from 40 mW to 240 mW, indicating that the emitting color of the sample is not meaningfully affected by the pump power.

 

Fig. 3 (a) UC spectra of the ZnWO₄: 0.05 mol% Er³⁺/0.5 mol% Yb³⁺ phosphor as a function of pump power at room temperature, and the inset of (a) shown the dependence of the green and red UC emission intensities on the pump power, (b) CIE chromaticity diagram of ZnWO₄: 0.05 mol% Er³⁺/0.5 mol% Yb³⁺ phosphor at different pump powers.

Download Full Size | PDF

3.2 Temperature sensing properties

It is known that the ²H_11/2 and ⁴S_3/2 levels of the Er³⁺ ion are thermally coupled and the transition ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 can be used for optical thermometry by using the FIR technique. The temperature sensing properties of the sample with x = 0.05 mol% was discussed attributing to the strongest emission of green light. From Fig. 4(a), it can be seen the peak positions have no change, but the intensity of the UC emission bands around 524 nm and 549 nm obviously varies with increasing the temperature of the sample. At low temperature about 83 K, the integrated intensity of band around 549 nm is greater than that about 524 nm while at higher temperatures it shows the reverse character. According to literatures [3, 9, 10, 29, 33–35], the FIR can be written as $\text{FIR}=\frac{I_{524}}{I_{549}}=Ae\text{xp}\left(\frac{-\Delta E}{kT}\right)$ Here, I₅₂₄ and I₅₄₉ are the integrated intensity of the ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 transitions, respectively. A is a constant, △E is the energy separation between the ²H_11/2 and ⁴S_3/2 levels, k is the Boltzmann constant, and T is absolute temperature.

 

Fig. 4 The temperature sensing properties of ZnWO₄: Er³⁺/Yb³⁺ phosphors, with x = 0.05 mol%: (a) Green UC emission spectra at different temperatures, (b) FIR relative to temperature and the inset of (b) is the sensor sensitivity as a function of temperature.

Download Full Size | PDF

Figure 4(b) shows a monolog plot of the FIR of green UC emissions at 524 nm and 549 nm as a function of absolute temperature in the range of 83–583 K. Clearly, the FIR value shows an upward trend with elevated temperature, reaching its maximum value at 583 K. By fitting the experimental data, the values of coefficient A and energy gap △E are about 19.304 and 788 cm⁻¹, respectively. The fitted value of 788 cm⁻¹ is close to the experiment value of 869 cm⁻¹, the latter was obtained from the UC spectra.

As is well-known, many factors should be taken into account to evaluate the practical application of optical temperature sensors, such as the sensor sensitivity, the measurement error, the operation temperature range and so on. The sensor sensitivity is a key parameter which is defined the rate of change of FIR in response to the variation of temperature. It can be mathematically expressed as follows [16, 29]:$\text{S}=\frac{\text{d}\left(\text{FIR}\right)}{\text{dT}}=\text{FIR}\frac{\Delta \text{E}}{{\text{kT}}^2}$, with which the sensor sensitivity was calculated and plotted in the inset of Fig. 4(b). It can be seen that the sensitivity keeps increasing with temperature and reaches its maximum value of about 0.0099 K⁻¹ at 583 K in our experimental temperature range. The optical temperature sensing ability of other Er³⁺ doped material is shown in Table 1. From the comparison, it is found that a highest sensitivity was obtained in ZnWO₄: Er³⁺/Yb³⁺ with wide temperature range. This property made ZnWO₄: Er³⁺/Yb³⁺ phosphor better candidate for high-performance optical thermometry within broader temperature range.

Table 1. The comparison of temperature sensing properties of Er doped oxides materials.

View Table | View all tables in this article

The measurement error △T also be estimated. According to the literature [36, 37], △T can be estimated from the sensitivity using the following equation: $\Delta \text{T}=\Delta \text{FIR}\left(\frac{{\text{kT}}^2}{\text{FIR*}\Delta \text{E}}\right)=\frac{\Delta \text{FIR}}{\text{S}}$ . Obviously, large sensitivity results in smaller error. The better accuracy can be still expected in the high temperature range using the UC emission of the ZnWO₄:Er³⁺/Yb³⁺ (x = 0.05 mol%) phosphor.

In general, luminescence of rare earth ion doped phosphors is very weak at high temperatures, due to intense thermal quenching. This phenomenon will lead to low signal-to-noise ratio and high measurement error. So, it is difficult to measure temperature change in the high temperature. From the Table 1, we can see that the temperature, at which the sensitivity maximum value is located, is highest. It indicated that ZnWO₄: Er³⁺/Yb³⁺ is promising for accurate thermometry applications even at high temperature.

3.3 Crystal structure

In order to characterize the structures of the prepared samples in detail, XRD Rietveld refinement was performed by using the GSAS-EXPGUI program [43,44]. Figure 5 shows the Rietveld plot and the schematic representation of the monoclinic ZnWO₄ unit cell for ZnWO₄:Er³⁺/Yb³⁺. The unit cell was modeled through a program called Visualization for Electronic and Structural Analysis (VESTA) using Rietveld refinement data. The structural refinement results are presented in Table 2. The coordinates of ZnWO₄ were used as an initial model where Er and Yb occupied the Zn site in wolframite cell. ZnWO₄:Er³⁺/Yb³⁺crystallizes in monoclinic structure [inset in Fig. 5(b)] with space group of P2/c (13). From Table 1, it can be seen that the lattice parameters a, b and c increase smoothly and unit cell volume (V) also gradually swells with x value increasing. These results suggest that Er³⁺/Yb³⁺ is successfully substituting Zn²⁺ in the lattice and that Er³⁺/Yb³⁺ is homogeneously incorporated into the lattice. Clearly, Zn-site substitution leads to monotonous increase in a, b, c, and unit cell volume, because both Er³⁺ and Yb³⁺ (1.004 Å [25] and 0.985 Å [45], respectively, 8 coordination number (CN)) are bigger in size than Zn²⁺ (0.74 Å, 8 CN) [46]. Meanwhile, as we can see, the Rietveld discrepancy factors are Rwp = 9.54% and Rp = 6.66% when x = 0.05 mol %, indicating the well agreement between the calculated of single phase ZnWO₄ and the measured pattern of low doped ZnWO₄ ceramics. When x = 0.2 mol %, the secondary phase began to appear. The Rietveld discrepancy factors are Rwp = 11.98% and Rp = 8.8%, respectively, which are still small though we refined it as a single phase. With the amount of Er³⁺ increased to 1.0 mol%, the result of refinement shows that the main phases were the monoclinic phase ZnWO₄ (Wt. = 53.44%) and the monoclinic phase Er₂WO₆/Yb₂WO₆ (Wt. = 46.56%). These results indicated the solid solubility limit is 0.2 mol% in ZnWO₄:Er³⁺/Yb³⁺ materials.

 

Fig. 5 Rietveld refinements and the structure for typical ZnWO₄: Er³⁺/Yb³⁺ samples, with x = 0.05 mol%, 0.20 mol% and 1.00 mol%, respectively.

Download Full Size | PDF

Table 2. Crystal data and structure refinement conditions for the ZnWO₄: Er³⁺/Yb³⁺ phosphors

View Table | View all tables in this article

The EDS spectra and element mappings of the ZnWO₄: 0.05 mol% Er³⁺/0.5 mol% Yb³⁺ sample were shown in Fig. 6. The EDS spectrum reveals evidently that the obtained phosphors mainly consisted of Zn, W, Er, Yb and O. From the element mappings, we can conclude that the compositions of Zn, W, Er, Yb and O were distributed over the whole particles. The results above suggest that Er³⁺ and Yb³⁺ ions were efficiently doped into the ZnWO₄ crystal lattice.

 

Fig. 6 EDS spectrum of the ZnWO₄: 0.05 mol% Er³⁺/0.5 mol% Yb³⁺ sample and the inset of Fig. 6 is elemental mappings.

Download Full Size | PDF

4. Conclusion

In summary, UC emissions from Er³⁺ ions in ZnWO₄: Er³⁺/Yb³⁺ phosphor were obtained. The obvious color-tunable upconversion emissions from green to red can be observed under infrared laser (980 nm) excitation. By the FIR technique, the optical temperature sensing properties was discussed using the ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 transitions of the Er³⁺ ions in the temperature range of 83–583 K. The maximum sensitivity of about 0.0099 K⁻¹ was obtained at 583 K in the process of UC. The result illustrates that an excellent optical temperature sensor with high sensitivity and wide temperature range can be designed based on the ZnWO₄: Er³⁺/Yb³⁺ phosphor.