1. Introduction

The upconversion (UC) luminescence properties of inorganic materials doped with rare earth (RE) ions have been investigated extensively due to their immense potential applications in the fields such as 3-dimentional displays, imaging of biological systems, temperature senators, etc [1–5]. Nowadays, the UC emissions covering wavelengths from violet to near infrared (NIR) have been achieved from RE doped materials. As a blue UC luminescence center couple, Tm³⁺/Yb³⁺ is frequently adopted, and the UC emission of Tm³⁺/Yb³⁺ has been achieved in different powdery host materials [6–8]. Among a large number of host materials, fluorides like NaYF₄ and oxides like Y₂O₃ have become the issues in research [9–12]. Comparing with these popularly investigated host materials, tungstates are high quality host materials due to their excellent optical properties, chemical and thermal stability.

The UC luminescence is well known as an anti-Stokes process in which low energy radiation, such as NIR or IR light, is converted to higher energy radiation like ultraviolet (UV) or visible light [13]. Whereas when the luminescent materials are exposed to the laser irradiation, a part of the absorbed photon energy in the materials is converted into heat energy through nonradiative process, and the material gets heated as a consequence. The optical heating could be not only used in medicine for local hypothermal treatment, but also for drilling nanoholes in organics and soft materials [14,15]. However, the internal heating induced by the laser irradiation has been reported rarely. S. B. Rai [16] and R. K. Verma [17] have studied on the laser induced optical heating in Er³⁺/Yb³⁺ codoped Gd₂O₃ nano-phosphor. and in Ho³⁺/Yb³⁺ codoped Ca₁₂Al₁₄O₃₃ phosphor, respectively. Moreover, the optical heating effects in Y₂O₃ phosphors doubly doped with Er³⁺/Yb³⁺ and Tm³⁺/Yb³⁺, triply doped with Eu³⁺/Er³⁺/Yb³⁺ have also been observed [18,19].

The laser irradiation can result in elevation of the sample temperature, and further bring temperature-relative influence on the spectroscopic properties of UC luminescence materials. For example, the cascade nonradiative relaxation rates will increase, the energy transfer rates between rare earth ions may change, and the dynamic processes for the upconversion emissions could also be varied. Therefore, the laser irradiation induced thermal effects are necessary to be investigated in order to comprehensively understand physical nature of UC luminescence processes under high power density excitation, which is also beneficial to the applications of different types of rare earth doped UC materials since the heat generation still accompanies the full UC luminescence process in any UC luminescence materials.

For the microwave heating, the microwave energy directly introduced into the chemical reactor, thus resulting in homogenous and rapid heating on the solution. In comparison with the traditional hydrothermal reaction, the microwave-assisted hydrothermal synthesis exhibits many advantages, such as short reaction-completed time, easy temperature and pressure control, quick and uniform heating, high reproducibility and so on. Therefore, in this study the microwave-assisted hydrothermal reaction was adopted for the sample preparation.

Here we reported on the preparation, structural and morphological characterization of Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures, as well as UC luminescence properties. The Er³⁺ doped NaY(WO₄)₂ microstructure was used as thermal probe to study on the laser irradiation induced thermal effect on Tm³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructures. It was found that the sample temperature was dependent on both the excitation power density and Tm³⁺ concentration. The time scanning processes indicated that UC intensities for Tm³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructures change slightly at low excitation power density, but change extremely at high excitation power density. The obtained results tell us that thermal effect evoked by the irradiation of high power IR laser could affect the analyses on the thermal sensing and UC luminescence mechanism, and the thermal effect should be carefully considered in the research work.

2. Experimental

2.1 Chemicals

The RE oxides Tm₂O₃ (purity 99.9%), Yb₂O₃ (purity 99.9%) and Y₂O₃ (purity 99.9%) were purchased from Shanghai Second Chemical Reagent Factory (China). Aqueous RE nitrate solutions were obtained by dissolving the corresponding RE nitrates with crystal water which was derived from a re-crystallization route. The detailed crystallization procedure can be found elsewhere [20]. Other chemicals like trisodium citrate (Na₃Cit) and Na₂WO₄•2H₂O purchased from Tianjin Reagent Chemicals Co. Ltd (China) were analytical grade regents and used directly without further purification.

2.2 Sample preparation

NaY(WO₄)₂:Tm³⁺/Yb³⁺ microstructures were prepared by a microwave-assisted hydrothermal method. The Yb³⁺ concentration was fixed to be 10 mol%, and Tm³⁺ concentrations were designed as 0.3, 0.5, 1, 1.5, 2 and 3 mol%, which present the molar percentages of the Y³⁺ replaced by Tm³⁺ (or Yb³⁺). The specific synthesis process of the target products is described as follows. Firstly, 2 mmol of Na₃Cit was dissolved in distilled water to form a 4 mL solution under magnetic stirring. The nitrate solutions Tm(NO₃)₃ and Yb(NO₃)₃ were mixed with designed stoichiometric ratios. Each mixed solution and 3 mL Y(NO₃)₃ solution containing 2 mmol Y(NO₃)₃ were dropped in the above Na₃Cit solution when Na₃Cit was totally dissolved. Stirring the mixed solution vigorously for 15 min. until white colloidal precipitate appeared. Then, 4 mL aqueous solution containing 4 mmol Na₂WO₄ was added under stirring for 20 min. The obtained mixed solution was transferred into a 30 mL silica glass vessel sealed with an aluminum cap provided by the manufacturer. After that, the vessel with precursor solution was placed inside a Microwave Synthesizer (Biotage Initiator, Sweden) and irradiated under microwave for 1 h at 180°C. The temperature and pressure of the reaction system could be monitored with internal-inserted system of the Microwave Synthesizer. As the microwave heating finished, the vessel was naturally cooled to room temperature, white precipitate appeared in the bottom of the vessel which was separated by centrifugation. After washed with distilled water and anhydrous ethanol for several times, the precipitate was dried at 75°C for 6 h, and went through a calcination process at 600 °C for 1 h to achieve final product NaY(WO₄)₂: Tm³⁺/Yb³⁺ microstructures. With this microwave hydrothermal technique a 1 mol% Er³⁺ and 10 mol% Yb³⁺ codoped NaY(WO₄)₂ microstructure which will be used as thermal probe was also prepared under the same conditions.

2.3 Characterization

The crystal structure of the samples was identified by X-ray diffraction (XRD) using a Rigaku D/MAX-Ultima⁺ diffractometer with graphite-monochromatized Cu-Kα radiation (λ = 0.15406 nm). In the measurement of XRD, the scanning region of 2θ angle was from 15° to 75°, and the scanning step size was 0.02°. The morphologies and particle sizes of the samples were characterized by a Hitachi S-4800 FE-SEM. The UC emission spectra and time scanning spectra under excitation of a 980 nm continuous fiber laser were measured by an F-4600 fluorescence spectrophotometer. The output power of fiber laser is linearly dependent on the working current. The excitation power density was estimated to be around 160W/cm² when the working current of fiber laser reached its maximum 2 A.

3. Results and discussion

3.1 Phase identification and morphologies

The XRD patterns for the NaY(WO₄)₂ microstructures with various Tm³⁺ concentrations and 10 mol% Yb³⁺ and for the one doped Er³⁺/Yb³⁺ were measured. It was found that all the samples exhibit very similar diffraction patterns as appeared in JCPDS card No.48-0886 for body-centered NaY(WO₄)₂ of tetragonal phase. As an example, the diffraction patterns for the samples doped with 1 and 3 mol% Tm³⁺ and 10 mol% Yb³⁺ are shown in Fig. 1 together with the pattern plotted by using the data reported in JCPDS card No. 48-0886. It can be found that in addition to the diffraction peaks from body-centered NaY(WO₄) ₂ there are no other peaks belonging to any impurities. This fact indicates that the Tm³⁺ doping does not obviously affect the crystal structure in the studied concentration region.

 

Fig. 1 X-ray patterns for NaY(WO₄)₂ microstructures doped with x mol%Tm³+/10 mol%Yb³⁺ (x = 1 and 3), and the standard diffraction pattern in JCPDS card No.48-0886

Download Full Size | PDF

The FE-SEM images for all the samples including the Er³⁺/Yb³⁺ codoped one were taken, and almost unchanged morphology was observed. This fact tells us that the Tm³⁺ doping does not affect the morphology of the final products, and the stable NaY(WO₄)₂ microstructures can be well reproducibly obtained via the microwave-assisted hydrothermal route. Figure 2(a) and 2(b) shows high and low magnification FE-SEM images for the NaY(WO₄)₂ microstructure doped with 3 mol% Tm³⁺ and 10 mol% Yb³⁺ as representatives. It can be seen that NaY(WO₄)₂ microstructures present sphere-shaped particles whose average size is around 2.5 μm, and each microsized particle is built by a large number of nanosheets with a mean thickness of about 30 nm.

 

Fig. 2 FE-SEM images for NaY(WO₄)₂ microstructure codoped with 3.0 mol% Tm³⁺/10 mol% Yb³⁺ (a) low magnification, (b) high magnification.

Download Full Size | PDF

3.2 Upconversion luminescence of Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microsctructures

The UC emission spectra for NaY(WO₄)₂ microsctructures with various Tm³⁺ concentrations and 10 mol% Yb³⁺ were measured under 980 nm laser with various output power. Intense blue and NIR emissions and weak red emission were observed in all the samples. As an example, Fig. 3 shows the UC emission spectra for NaY(WO₄)₂:1.0%Tm³⁺/10%Yb³⁺ microstructures measured under 980 nm laser when the excitation power density changed from 19 to 160 W/cm². It should be mentioned that the area of the laser beam spot maintained as a constant. More importantly, each UC spectrum was measured just at the very beginning as the laser just conducted to the sample, and the time interval between two spectral measurements was set to be long enough time during which the laser was stopped, thus the heating effect induced by the laser irradiation is expected be avoided. In Fig. 3 the three emission bands centered at around 476, 649 and 796 nm can be assigned, respectively, to the ¹G₄→³H_6, ¹G₄→³F_4, and ³H₄→³H₆ transitions of Tm³⁺ [21]. Meanwhile, their intensities increase with the increase of excitation power density.

 

Fig. 3 Excitation-power-density-dependent up-conversion emission spectra of NaY(WO₄)₂:1.0% Tm³⁺ /10%Yb³⁺ under 980 nm excitation.

Download Full Size | PDF

The integrated emission intensities for the samples with different Tm³⁺ doping concentrations were derived by numeric analyses on the above UC emission spectra, but the red emission was not taken into account since its intensity is too weak which can result in large calculation errors. For a multi-photon UC process, the UC luminescence intensity $I_{up}$depends on the pump laser power $P$, which can be mathematically expressed as,

 

Fig. 4 Dependences of blue and near infrared integrated UC emission intensities on the excitation power density for the samples doped with 1 mol% Tm³⁺ (a) and 3 mol% Tm³⁺ (b).

Download Full Size | PDF

3.3 Temperature sensing calibration of Er³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructure

As mentioned above, the sample temperature will be elevated when the sample is exposed to the high power density laser. Comprehensive understanding on the laser-irradiation-reduced thermal effect is very important since the sample temperature change will result in the variations of multiphonon cascade nonradiative relaxation rates and energy transfer rates. Moreover, laser irradiation induced thermal effect will directly affect the analysis on the UC processes. In other aspect the laser heating effect in rare earth nanomaterials may find applications in opto-thermal therapy in biomedicine field. However, the temperature measurements for the Tm³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructures are difficult to be carried out, since the heat energy generated by the laser irradiation is very less and mainly localized in the area on the sample where the laser beam spot exposed, and that the temperature will change if any sensor gets in contact with the laser beam spot [25–28]. Therefore, the temperature detection for sample under laser irradiation is a challenge to the researchers.

In order to solve the problem with the temperature detection in Tm³⁺/Yb³⁺ codoped microstructures, the 1 mol% Er³⁺/ 10 mol% Yb³⁺ doped NaY(WO₄)₂ microstructure is proposed to be adopted as thermal probe to confirm the sample temperature via UC spectral measurements. In doing so, the Er³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructure was uniformly mixed with the Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures with various Tm³⁺ concentrations. The mixing weight ratio of Er³⁺/Yb³⁺ doped NaY(WO₄)₂ sample to Tm³⁺/Yb³⁺ doped samples was the same as 4:100 in all the cases for the Tm³⁺/Yb³⁺ doped samples with various Tm³⁺ concentrations. It can be believed that the influence of Er³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructure on the mixture temperature can be neglected when the mixture is irradiated under 980 nm laser, since the content of Er³⁺/Yb³⁺ doped NaY(WO₄)₂ microstructure is less.

The temperature sensing performance of Er³⁺ is decided by its special energy level configuration [27,28]. The energy gap between the ²H_11/2 and ⁴S_3/2 levels of Er³⁺ is about 850 cm⁻¹ in most inorganic hosts [29]. These two green emitting levels of Er³⁺ ion are in thermally coupled regime. The population of thermally coupled ²H_11/2 and ⁴S_3/2 levels follows Boltzmann’s distribution law. Thus, the ratio $R$ of fluorescence intensities for the transitions from the thermally coupled levels ²H_11/2 and ⁴S_3/2 to the ground state ⁴I_15/2 could be described as [30–32]

 

Fig. 5 Dependence of Er³⁺ fluorescence intensity ratio on the sample temperature (squared dots); numerical fitting curve by using the theoretical formula $R=C\exp (-\Delta E/kT)$ (solid line). The insert (a) and (b) show the excitation and emission spectrum for Er³⁺/Yb³⁺ co-doped microstructure measured at room temperature.

Download Full Size | PDF

3.4 Thermal effect induced by 980 nm laser irradiation

The idea for studying on the thermal effect induced by 980 nm laser irradiation in Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures is already described in section 3.3. In order to check the feasibility of applying this temperature sensing technique, the UC emission spectra for all the mixtures containing Er³⁺/Yb³⁺ codoped sample and Tm³⁺/Yb³⁺ codoped samples with different Tm³⁺ concentrations were measured under 980 nm laser excitation and are shown in Fig. 6(a). The excitation power density was fixed to be 112 W/cm² in the measurements of UC spectra. It should also be mentioned that each spectrum was measured at the beginning when the 980 nm laser beam just irradiated at the sample, thus the thermal effect induced by laser irradiation may be avoided as can as possible.

 

Fig. 6 (a) UC emission spectra for the mixture doped with various Tm³⁺ concentrations under 980 nm excitation; (b) normalized UC spectra plotted by using the data in Fig. 6(a).

Download Full Size | PDF

From Fig. 6(a) it can be seen that the spectra include the green UC emissions corresponding to ²H_11/2/⁴S_3/2→⁴I_13/2 transitions from Er³⁺ in addition to the intrinsic emissions peaking at 476, 649 and 796 nm corresponding to ¹G₄→³H_6, ¹G₄→³F_4, and ³H₄→³H₆ transitions of Tm³⁺. The green UC emission peaks can well differentiate from Tm³⁺ emission peaks without any spectral overlap, thus indicating the Er³⁺/Yb³⁺ codoped microstructure may be a good thermal probe for studying on the heating effect of 980 nm laser irradiation in Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures. It is found from Fig. 6 (a) that the UC luminescence intensities for both the blue (476 nm from Tm³⁺) and NIR (796 nm from Tm³⁺) emissions increase first with increasing Tm³⁺ concentration and reach their maximums at 2 mol% of Tm³⁺, and then decrease with the continued increase of Tm³⁺ concentration. The concentration quenching of Tm³⁺ UC luminescence is caused by the cross relaxation between Tm³⁺ ions [25]. It should also be noted from the excitation spectrum that (insert (a) of Fig. 5) there is no effective excitation at 476 nm which matches the emission of Tm³⁺, moreover the content of Er³⁺/Yb³⁺ doped sample in the mixture is less. Therefore, the radiation re-absorption process (the 476 nm emission of Tm³⁺ is absorbed by Er³⁺, sometimes also regarding as energy transfer process, but here we consider it as radiation re-absorption process) can be neglected. Furthermore, the radiation re-absorption rate is usually much smaller than the energy transfer rate based on the electric multipole interaction since the energy transfer takes place between luminescent centers in one particle and the distance between the luminescent centers can be very short, nevertheless the energy transfer between Tm³⁺ and Er³⁺ will not happen since electromagnetic interaction between Tm³⁺ and Er³⁺ cannot take effect in different particles.

Figure 6(b) shows the UC emission spectra normalized to ²H_11/2→⁴I_13/2 (Er³⁺) intensity plotted by using the data in Fig. 6 (a). It is found that two green emissions can be well identified, thus indicating the Er³⁺/Yb³⁺ codoped microstructure is applicable for temperature sensing, though its content mixed into the Tm³⁺/Yb³⁺ codoped microstructures is less, the spectral intensities seem intense enough for the numerical calculations. It should also be noted that when the doping concentration is lower than 2.0 mol%, the fluorescence intensity ratio $R$ for the two green UC emissions of Er³⁺ is almost the same, but the $R$ increases with increasing the Tm³⁺ concentration. This fact indicates that for the sample with higher Tm³⁺ concentration the temperature gets elevated at very beginning when the 980 nm laser just irradiates on the sample. This means the sample with higher Tm³⁺ exhibits more obvious laser irradiation heating effect.

As we have mentioned above, to further study on the thermal effect of laser irradiation in Tm³⁺/Yb³⁺ codoped microstructures, the UC emission spectra for the each mixture of Er³⁺/Yb³⁺ doped sample and Tm³⁺/Yb³⁺ doped sample with certain Tm³⁺ concentration was measured under excitation of 980 nm laser at different excitation power density. It should be stated that each spectrum was measured immediately when the sample was just exposed to the 980 nm laser. The scanning speed for the spectral measurements was set to be 2400 nm/s. Moreover, the time interval between two spectral measurements was long enough, and the laser switched off during the idle time of spectral measurements, so that the durative heating effect on the sample can be neglected. The UC spectra include both the emissions from Er³⁺ and Tm³⁺ as analogously observed in Fig. 6(a). The green UC emissions of Er³⁺ were used to probe the sample temperature by taking the green emission intensity ratio into Eq. (2). The temperatures for all the samples with various Tm³⁺ concentration excited by 980 nm laser with various output power densities and constant laser beam area were calculated. Figure 7 shows the dependences of the sample temperature on the Tm³⁺ concentration and excitation power density. It can be seen from Fig. 7 that for each sample with certain Tm³⁺ concentration the sample temperature increases with increasing excitation power density. This result can be easily understood, for most photoluminescence materials the excitation energy cannot be fully converted to the light irradiations, but partial excitation photon energy becomes heat energy making the sample temperature elevated. If the excitation energy is higher, the converted heat energy is more, and then the sample temperature would be higher. From Fig. 7 it can also be found that when the excitation power density keeps constant, the temperature for the samples with Tm³⁺ concentration lower than 1.5 mol% increases slightly, but it increases greatly with increasing Tm³⁺ concentration when the doping concentration is higher than 1.5 mol%. It was already found from Fig. 6(a) that the maximum UC intensity was achieved in the sample with 2 mol% Tm³⁺, and that the UC intensity increases with increasing Tm³⁺ concentration when Tm³⁺ concentration is lower than 2 mol% and the excitation power density keeps unchanged. This fact means that when the Tm³⁺ doping concentration is lower than 2 mol%, the increased excitation energy absorbed by Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures caused by increasing Tm³⁺ concentration mainly contributes to the UC photon emissions, therefore, the sample temperature increases not obviously as seen in Fig. 7. When the Tm³⁺ concentration is higher than 2 mol%, the excitation energy transferred from Yb³⁺ would be more, nevertheless the UC luminescence is quenched via phonon assisted energy transfers in which the phonon generation processes are dominant, thus the generated heat energy would be high, thus resulting in sample temperature elevation.

 

Fig. 7 Dependences of sample temperature on excitation power density and Tm³⁺ doping concentration.

Download Full Size | PDF

To further understand the thermal effect of laser irradiation in NaY(WO₄)₂:Tm³⁺/Yb³⁺ phosphors, the time scanning UC emission spectra for varied excitation power densities were measured by monitoring the blue (475 nm) emission when the samples were continuously excited by 980 nm laser. It should be noted that during the whole measurement there was no chemical or physical damage on the samples due to their high chemical and physical stability resulting from the sample’s post calcination at 600 °C for 1 h. Meanwhile, the time scanning processes can be re-produced, thus indicating the irradiation damage on the samples does not take effect. The open circled dots in Fig. 8 (a) and 8(b) show the time scanning spectra for the samples doped with 3 and 0.3 mol% Tm³⁺ measured under the same experimental conditions. It should be mentioned that each measurement is started until the sample is cooled to room temperature. It is found that when the excitation power density is lower than 112 W/cm², the UC luminescence intensities for both the samples with 0.3 and 3 mol% Tm³⁺ nearly not change with increasing irradiation time. However, when the excitation power density is higher (for example, 160W/cm²), the UC luminescence intensities decrease dramatically with increasing irradiation time. The decrease of UC luminescence intensity with irradiation time reflects the elevation of sample temperature, since the multiphonon cascade nonradiative relaxation of luminescence level increases with increasing the sample temperature [33]. It should be noted that when the excitation power density is 160 W/cm², the UC intensities for both the samples are still decreasing within 30 minutes and draw an approximate linear relationship toward the irradiation time. The data for higher excitation power density in Fig. 8 (a) and 8(b) were fit to the formula $y=a+kx$, and the $k$ values, which is defined as UC luminescence intensity change rate, were found from the fitting to be 0.19 and 0.44 Ins/s (here Ins presents the relative luminescent intensity, s means time unit, namely second) for the samples with 3 and 0.3 mol% Tm³⁺, respectively. From the above analyses on the results in Fig. 6 it can be deduced that the temperature of the sample with lower Tm³⁺ concentration would be low and increases with irradiation time should also be slow. However, the change rate of UC luminescence intensity for the sample with low Tm³⁺ concentration is large. This is due to the fact that the nonlinear dependence of luminescence intensity on the temperature in quenching process, usually, follows the Arrhenius’s model, in which the relation between the luminescence intensity and temperature can be expresses as

 

Fig. 8 Time scanning spectra for blue UC emission of NaY(WO₄)₂ microstructurs doped with 3 mol%Tm³⁺/10 mol%Yb³⁺ (a) and 0.3 mol%Tm³⁺/10 mol%Yb³⁺ (b) measured at excitation power density under excitation of 980 nm laser.

Download Full Size | PDF

4. Conclusion

The NaY(WO₄)₂ microstructures with various Tm³⁺ concentrations and 10 mol% Yb³⁺ were successfully synthesized by a microwave-assisted hydrothermal method. 1 mol% Er³⁺/10 mol% Yb³⁺ codoped NaY(WO₄)₂ microstructure derived from the same synthesis technique under the same experimental conditions was used as temperature sensing probe to explore the thermal effect induced by 980 nm laser irradiation in Tm³⁺/Yb³⁺ codoped NaY(WO₄)₂ microstructures. It was confirmed that the target products existed in tetragonal phase and nanosheets-built microsphere-like shape. It was found that the sample temperature depended on both the excitation power density and the Tm³⁺ concentration: the sample temperature at the beginning of laser irradiation increased with increasing the excitation power density and the Tm³⁺ concentration. Moreover, time scanning spectra for the samples with lower and higher Tm³⁺ concentrations showed that the UC luminescence intensity did not change with irradiation time when the excitation power density was lower, but decreased greatly when excitation power density was higher, thus reflecting the sample temperature change with laser irradiation time which was explained by Arrhenius’s model for temperature quenching.