*2.2. Characterization*

Powder X-ray di ffraction (XRD) studies were carried out on PANalytical X'Pert Pro di ffractometer equipped with Anton Paar TCU 1000 N Temperature Control Unit using Ni-filtered Cu *K*α radiation (*V* = 40 kV, *I* = 30 mA).

Transmission electron microscope images were taken using transmission electron microscopy (TEM)PhilipsCM-20SuperTwinwith160kVofacceleratingvoltageand0.25nmofopticalresolution.

The hydrodynamic size of the nanoparticles was determined by dynamic light scattering (DLS), conducted in Malvern ZetaSizer at room temperature in polystyrene cuvette, using distilled water as a dispersant.

The emission spectra were measured using the 266 nm excitation line from a laser diode (LD) and a Silver-Nova Super Range TEC Spectrometer form Stellarnet (1 nm spectral resolution) as a detector. The temperature of the sample was controlled using a THMS600 heating stage from Linkam (0.1 1C temperature stability and 0.1 1C set point resolution).

Luminescence decay profiles were recorded using FLS980 Fluorescence Spectrometer from Edinburgh Instruments with μFlash lamp as an excitation source and R928P side window photomultiplier tube from Hamamatsu as a detector.

#### **3. Results and Discussion**

The yttrium aluminum/gallium garnets crystallize in a cubic structure of Ia3d space group. The general formula of garnets is expressed as follows: A3B2C3O12, where three di fferent metallic sites are represented by dodecahedral site (A), octahedral site (B) and tetrahedral site (C), which in our case are occupied by eight-fold coordinated Y3+ ions, six-fold coordinated Al<sup>3</sup>+/Ga3+ ions and four-fold coordinated Al<sup>3</sup>+/Ga3+ ions, respectively. The optically active ions introduced to the structure may occupy di fferent crystallographic sites, which results from the similarities in the coordination number, ionic radii and ionic charge between the host and dopant metal. Therefore, lanthanides (Ln<sup>3</sup>+) prefer to replace A site, while (TM) mainly substitute B and C sites. Additionally, depending on the size of TM ion, they occupy larger (B) (ionic radii 0.67 Å for Al3+ and 0.76 Å for Ga3+) or smaller (C) (0.53 Å for Al3+and 0.61 Å for Ga3+) metallic sites. An XRD analysis was used to verify the phase purity of synthesized materials. It is evident that the obtained di ffraction peaks of V-doped Y3Al5−<sup>x</sup>GaxO12 nanocrystals correspond to the reference patterns of cubic structures of adequate host materials (Figure 1a). Observed peaks broadening can be assigned to the small size of the nanoparticles. The *a* cell parameter increases linearly as the Ga3+-dopant concentration increased, which results from the enlargement of the crystallographic cell associated with the di fference in the ionic radii of Al3+and Ga3+ ions (rAl<sup>3</sup>+ < rGa3+) (Figure 1b). However, it was found that Ga3+ ions preferentially occupy four-fold coordinated sites of Al3+ rather than the octahedral counterpart. This phenomenon can be explained based on the stronger covalency of Ga3<sup>+</sup>-O2− bonds with respect to the Al3<sup>+</sup>-O2− ones and the lowering of repulsive forces between cations, providing stabilization of the crystal structure [31,32]. On the other hand, the slight shift of the XRD peaks with respect to the reference pattern arises from the implementation of V ions into Y3Al5−<sup>x</sup>GaxO12 lattice. It was found that Y3Al5−<sup>x</sup>GaxO12 matrix is a suitable host material for three di fferent V oxidation states, namely V3<sup>+</sup> and V5<sup>+</sup> [23,26,33]. The replacement of Ga3+ and Al3+ ions by V ions is possible due to their comparable ionic radii, which in the case of four-fold coordinated V5<sup>+</sup> and V3<sup>+</sup> ions are 0.54 Å, 0.64 Å, respectively, and for six-fold coordinated <sup>V</sup>5+,V4<sup>+</sup> and V3<sup>+</sup> ions are 0.68 Å, 0.72 Å and 0.78 Å, respectively. As can be seen from the TEM images, synthesized powders consist of well-crystalized and highly agglomerated nanocrystals (Figure 1c,e,g,i,k). The hydrodynamic sizes of the aggregates of Y3Al5−<sup>x</sup>GaxO12 nanocrystals examined using DLS analysis were found to be around 300 nm (Figure 1d,f,h,j,l).

**Figure 1.** (**a**) XRD patterns of Y3Al5−<sup>x</sup>GaxO12 nanocrystals, doped with 0.1% V; (**b**) influence of the Ga3+ concentration on the *a* cell parameter; (**<sup>c</sup>**,**e**,**g**,**i**,**k**): the morphology of Y3Al4GaO12, Y3Al3Ga2O12, Y3Al2Ga3O12, Y3AlGa4O12, Y3Ga5O12, respectively; (**d**,**f**,**h**,**j**,**l**): the distribution of the hydrodynamic size of aggregates.

Luminescent properties of V- doped Y3Al5−<sup>x</sup>GaxO12 nanocrystals were investigated upon 266 nm of excitation in the −150 ◦C to 300 ◦C (123.15 K to 573.15 K) temperature range (Figure 2a). The emission spectrum obtained at −150 ◦C consists of three transition bands, for materials with Ga3+ concentration from 1 to 4, and of two emission bands for YGG, being related to the presence of different V oxidations states - <sup>V</sup>5+, V4<sup>+</sup> and V3+. In the course of our previous investigation, it was found that due to the difference in the ionic radii and the charge, V5<sup>+</sup> ions preferentially occupy surface sites of Al3+, while V3<sup>+</sup> and V4<sup>+</sup> are mainly located in the core part of the nanoparticles [26,33]. The first broad emission band at 520 nm is attributed to the charge transfer transition of V5<sup>+</sup>(V<sup>4</sup><sup>+</sup> → <sup>O</sup>2−). The second band at 640 nm originates from 2E → 2T2 radiative transition of V4<sup>+</sup> ions, while the band at 820 nm is associated with 1E2 → 3T1g transition of V3<sup>+</sup> ions. As can be seen, the addition of Ga3+ ions significantly affects the luminescent properties of Y3Al5−<sup>x</sup>GaxO12:V nanocrystals (Figure 2b). The presented results stay in agreemen<sup>t</sup> with the observations obtained for the vanadium doped yttrium aluminum oxide and lanthanum gallium oxide nanoparticles [23,26]. The representative emission spectra measured at −150 ◦C indicate that the increase of Ga3+ concentration caused the enhancement of the V4<sup>+</sup> emission intensity in respect to the V5<sup>+</sup> and V3<sup>+</sup> ones. This effect results from the large ionic radii of <sup>V</sup>4+, which significantly exceeds Al3+ ones. Therefore, V4<sup>+</sup> cannot efficiently replace Al3+ in the structure. However, when the concentration of Ga3+ ions gradually increases, the number of the crystallographic sites that can be occupied by V4<sup>+</sup> rises up, leading to the enhancement of 2E → 2T2 emission intensity. Moreover, the Ga-doping induces the reduction of the distance between V4<sup>+</sup> and V5<sup>+</sup> ions facilitating the energy transfer between them, which contributed to the V4<sup>+</sup> luminescent intensity increase. It is worth noticing that the emission of trivalent V dominates in the spectrum up to x = 4, while in the case of YGG <sup>V</sup>4+, the emission band prevails. To quantify these changes the histogram presenting the contribution of the emission intensities (calculated as an integral emission intensity in appropriate spectral range) of particular oxidation state of vanadium ions to the overall emission intensity as a function of Ga3+ concentration is presented in Figure 2c. The observed enhancement of V4<sup>+</sup> emission intensity with respect to the V5<sup>+</sup> with an increase of Ga3+ concentration causes tuning of the emission color toward red emission (Figure 2d). However, for YGG:V, orange emission was found. As has been already proven, the V5<sup>+</sup> ions are located mainly in the surface part of the nanocrystals [23]. Since the morphology and the size of the nanoparticle is independent on the Ga3+ concentration, the number of V5<sup>+</sup> can be assumed to be constant. The confirmation of this hypothesis is the fact that its lifetime (<sup>&</sup>lt;τV5+<sup>&</sup>gt; = 6.4 ms) is independent on the host stoichiometry (Figure S1). On the other hand, the average lifetime of V3<sup>+</sup> and V4<sup>+</sup> shortens consequently from 7.6 ms to 7.0 ms and 1.2 ms to 0.5 ms, respectively, with Ga3+ concentration (x changed from 1 to 5).

**Figure 2.** (**a**) The energy diagram of V ions at different oxidation states; (**b**) the influence of Ga-doping on the V emission spectrum (at −150 ◦C under 266 nm) in Y3Al5−<sup>x</sup>GaxO12 nanomaterials at 0 ◦C; (**c**) the contribution of emission intensity of particular oxidation state of V ions into the overall emission spectrum of V-doped Y3Al5−<sup>x</sup>GaxO12 nanocrystals; (**d**) the Commission internationale de l'éclairage CIE 1931 chromatic coordinates calculated for V:Y3Al5−<sup>x</sup>GaxO12 nanocrystals at 0 ◦C.

In order to evaluate how the spectral changes of Y3Al5−<sup>x</sup>GaxO12 nanocrystal, induced by the stoichiometry modification, affect the performance of analyzed nanoparticles for noncontact temperature sensing, their luminescence spectra were analyzed in a wide range of temperature (from −150 ◦C to 300 ◦C) (Figure 3a, Figure S2). In the course of these studies, it was found that emission intensity of each V ion is quenched by temperature; however, their luminescence thermal quenching rates differ (Figure 3b–d). In the case of <sup>V</sup>5+, emission intensity is gradually quenched by almost two orders of magnitude with temperature. However, correlation between Ga3+ introduction and temperature of thermal quenching was not observed. This effect is understandable, since, as has been shown before, V5<sup>+</sup> occupy mainly surface part of the nanoparticles. In turn, the emission intensity of V4<sup>+</sup> initially decreases with temperature and above some critical temperature, it significantly increases as the temperature grows, which results from the efficient V5<sup>+</sup> → V4<sup>+</sup> energy transfer. It was found that the threshold temperature above which rise up of intensity was observed lowers with Ga3+ concentration (from around 10 ◦C for Y3Al4GaO12 to −100 ◦C for Y3AlGa4O12 and YGG). Additionally the magnitude of the intensity increase growths with Ga3+ content. This phenomenon can be explained by the increase of the V5<sup>+</sup> → V4<sup>+</sup> energy transfer probability. Higher numbers of Ga3+ sites in the structures promote the stabilization of the V4<sup>+</sup> ions, which, as a consequence, shortens the average distance between V5<sup>+</sup> and V4<sup>+</sup> facilitating interionic interactions. Due to the fact that energy of V5<sup>+</sup> excited state is higher than that of <sup>V</sup>4+, the energy transfer between them occurs with the assistance of the phonon. According to the Miyakava-Dexter theory, the probability of this process is strongly dependent on temperature, which is in agreemen<sup>t</sup> with our data [34]. It needs to be noted that although V5<sup>+</sup> ions serve as a sensitizers for <sup>V</sup>4+, there is no correlation between Ga3+ concentration and the V5<sup>+</sup> luminescence thermal quenching. This comes from the fact that in the case of V5<sup>+</sup> intensity the luminescence thermal quenching process plays dominant role over V5<sup>+</sup> → V4<sup>+</sup> energy transfer. The correlation between Ga3+ concentration and the luminescent thermal quenching rate is also evident in the case of V3<sup>+</sup> ions. The higher the amount of Ga3+, the lower the thermal quenching rate of the 1E2 → 3T1g emission band. Above 100 ◦C, the V4<sup>+</sup> emission intensity becomes so efficient that its intensity dominates over the V3<sup>+</sup> ones and thus hinders its emission intensity analysis. In the case of YGG, the V3<sup>+</sup> emission is impossible to detect.

Since the emission intensity of V ions in Y3Al5−<sup>x</sup>GaxO12 nanocrystals is strongly affected by the temperature changes, a quantitative analysis, which verify their performance for non-contact temperature sensing, was performed. For this purpose, the relative sensitivities (*S*) of three different intensity-based luminescent thermometers were calculated according to the following Equation (1):

$$S = \frac{1}{\Omega} \frac{\Delta \Omega}{\Delta T} \cdot 100\% \,\tag{1}$$

where Ω corresponds to the temperature dependent spectroscopic parameter, which in this case is represented by emission of adequate V ions (S1 for <sup>V</sup>5+, S2 for V4<sup>+</sup> and S3 for V3<sup>+</sup>), and ΔΩ and Δ*T* indicate to the change of Ω and temperature, respectively.

The maximal values of relative sensitivity (S1) of V5<sup>+</sup>-based luminescent thermometer, which exceed 2%/◦C, were found at temperatures below −100 ◦C and with increase of temperature S1 gradually decreases reaching 1.34%/◦C, 1.12%/◦C, 1.13%/◦C, 1.30%/◦C and 0.76%/◦C for Y3Al4GaO12, Y3Al3Ga2O12, Y3Al2Ga3O12, Y3AlGa4O12 and Y3Ga5O12, respectively, in the biological temperature range (0 ◦C–50 ◦C). The highest value of the S1 was found at −150 ◦C for Y3Ga5O12, which is in agreemen<sup>t</sup> with our expectation that short distance between V5<sup>+</sup> and V4<sup>+</sup> facilitates the interionic energy transfer between them. The presented correlations confirm that relative sensitivity of temperature sensors based on V5<sup>+</sup> emission intensity can be modulated by varying the Ga3<sup>+</sup>-concentration (Figure 3e). In case of Y3Al5−<sup>x</sup>GaxO12:V<sup>4</sup><sup>+</sup> temperature sensors, the highest value of sensitivity reveal the YGG nanocrystals (S2max = 1.34%/◦C at −15 ◦C), and its value gradually decreases with the lowering of Ga3+ concentration. Moreover, the temperature at which maximal S2 was found decreases with Ga3+ concentration from 75 ◦C for Y3Al4GaO12 to −15 ◦C for YGG. This phenomenon is also observed in the case of biological

temperature range, where reducing the Ga3+ concentration the S value decreases from 1.32%/◦C at 0 ◦C to 0.2%/◦C at 30 ◦C for Y3Ga5O12 to Y3Al4GaO12 (Figure 3f). It should be mentioned here that usable temperature range for this luminescent thermometer (temperature range in which Ω reveals monotonic change) is limited, and the most narrow one was found for YGG (from −100 ◦C to 120 ◦C). The negative values of S2 come from the fact of the intensity trend reversal. Hence, the balance between relative sensitivity and the usable temperature range can be optimized by the appropriate host material composition. Therefore, depending on the type of application of such luminescent thermometer, including required relative sensitivity and operating temperatures range, different stoichiometry of host material can be proposed. Since the V3<sup>+</sup> emission intensity monotonically decreases in the temperature range below 200 ◦C the relative sensitivity S3 reveals positive values with the single maxima at temperature which is dependent on the Ga3+ concentration (Figure 3g). The increase of Ga3+ amount causes the reduction of both value of the S3 and the temperature of S max from 1.08%/◦C at 152 ◦C for Y3Al4GaO12 to 0.45%/◦C at 51 ◦C for Y3AlGa4O12.

**Figure 3.** (**a**) Thermal evolution of emission spectrum of Y3AlGa4O12:V nanocrystals; (**b**–**d**) the influence of local temperature on the emission intensity of <sup>V</sup>5+, V4<sup>+</sup> and <sup>V</sup>3+, respectively; (**<sup>e</sup>**–**g**) corresponding relative sensitivities.

Although the performance of the intensity-based luminescent thermometer, which take advantage from <sup>V</sup>5+, V4<sup>+</sup> and V3<sup>+</sup> emission, are very promising, the reliability of accurate temperature readout is limited due to the fact that emission intensity of a single band may be affected by the number of experimental and physical parameters. Therefore, most of the studies concern the bandshape luminescent thermometer, for which relative emission intensity of two bands is used for temperature sensing. Taking advantage of the fact that emission intensities of V5<sup>+</sup> and V4<sup>+</sup> ions reveal opposite temperature dependence, their luminescence intensity ratio (*LIR*) can be used as a sensitive thermometric parameter:

$$LIR = \frac{\text{V}^{5+} (\text{V}^{4+} \to \text{O}^{2-})}{\text{V}^{4+} (\text{"E} \to \text{\${}\_2\text{T}\_2\text{}})} , \tag{2}$$

Analysis of the thermal evolution of *LIR* reveals that for each stoichiometry of the host material the decrease of *LIR*'s value by over three orders of magnitude can be found for −150–300 ◦C temperature range (Figure 4a). Observed thermal changes of *LIR* significantly exceed those noticed for single ion emission. The relative sensitivities of LIR-based luminescent thermometers (*S4*) were defined as follows:

$$S\_4 = \frac{1}{LIR} \cdot \frac{\Delta LIR}{\Delta T} \cdot 100\%\_{\prime} \tag{3}$$

Thereby, the relative sensitivities calculated for LIR-based luminescent thermometers reached values that exceed 2%/◦C (Figure 4b). Thermal evolution of *S4* attains single maxima at temperature TSmax. As was shown before, both the *S4max* and TSmax can be successfully modified by the incorporation of the Ga3+ ions. The increase of the Ga3+ concentration causes the lowering of the TSmax from 20 ◦C for Y3Al4GaO12 to −100 ◦C for Y3Ga5O12, while the maximal relative sensitivity increases from 1.47%/◦C

for Y3Al4GaO12 to 2.48%/◦C for Y3Ga5O12 (Figure 4c,d). However, the maximal value of *S4* = 2.64%/◦C was found for Y3Al2Ga3O12. It needs to be mentioned here that, to the best of our knowledge, described nanocrystals reveal the highest values of relative sensitivity for vanadium-based luminescent thermometers up to date. Moreover, it was found that the higher the Ga3+ content (Y3Al4GaO12- Y3AlGa4O12), the more significant the change of CIE 1931 chromatic coordinates is (Figure 4e,f).

**Figure 4.** (**a**) Thermal evolution of luminescence intensity ratio (LIR); (**b**) their relative sensitivities for Y3Al5−<sup>x</sup>GaxO12 nanocrystals; (**c**) the temperature at which the maximal value of S4 was observed; (**d**) S4max as a function of Ga3+ concentration; (**<sup>e</sup>**,**f**) the CIE 1931 chromatic coordinates calculated for Y3Al4GaO12:V and Y3AlGa4O12:V nanocrystals, respectively.
