Optical Temperature Sensor Capabilities of the Green Upconverted Luminescence of Er³⁺ in La₃NbO₇ Ceramic Powders

Abstract

We present a study of the Er³⁺ upconverted luminescence in erbium doped Lanthanum Niobium Oxide, La₃NbO₇, ceramic powder, prepared by solid state reaction. This study focuses on the analysis of the feasibility of this system as a temperature sensor. Efficient UC luminescence was observed under the 975 nm excitation showing intense green, red and NIR (850 nm) emission bands. The NIR luminescence centred at about 850 nm and lying on the first biological window is mainly insensitive to the temperature. In contrast, the upconverted green bands, associated with the (²H_11/2,⁴S_3/2) →⁴I_15/2 transitions, showed a high sensibility to temperature. Their temperature dependence was studied from RT up to 525 K, paying special attention to the physiological range of temperature (303–318 K). The high thermal sensitivities obtained, in comparison with other Er³⁺ and Er³⁺-Yb³⁺ based optical temperature sensors in such ranges, suggest the potential application of this phosphor in thermal sensing, suitable for both biological systems and other industrial applications requiring higher temperatures.

1. Introduction

Measuring temperature accurately and reliably is very important in several application areas, ranging from health care to industrial applications [1,2,3,4,5]. The working mechanism of optical temperature sensors is based on the temperature dependence of the material’s optical properties, which allows for the monitoring of temperature in certain locations such as in electrical power stations, refineries, and coal mines. These sensors can also be useful in building fire detection [6,7]. In addition to precision and reliability, other characteristics such as high sensitivity and resolution and fast response are also required when using these materials as temperature sensors [2]. Changes in optical properties and their dependence on temperature have been studied in different ways: the spectral position of emission and absorption lines, their intensity, bandwidths and decay time and even polarization [2]. Since the discussions in the 1970s of the possible application of the fluorescence intensity ratio (FIR) technique for temperature sensing[8], this technique is among the most widely used due to its excellent advantages in terms of sensitivity and stability [9]. The classical use of this technique requires the presence of thermally coupled levels (TCL), two nearby energy levels which are assumed to be in a state of thermal quasi-equilibrium [10], in which luminescence intensities are measured. The energy distance between these TCLs must be small enough to allow a redistribution of the population of thermal origin between the lower and higher energy levels, which is the origin of the FIR technique. However, at the same time there has to be a separation between levels large enough to solve the emission lines originating in each one of them. A typical value between 200 and 2000 cm⁻¹ [11] is usually accepted as the energy gap between levels that allows the application of the FIR technique. Most of the trivalent lanthanide ions (Pr³⁺, Nd³⁺, Eu³⁺, Ho³⁺, Er³⁺, Tm³⁺) fulfil the above requirement and their introduction in small amounts into host materials of different types allows for the use of these materials as potential optical thermometers [12,13,14,15,16]. The narrow absorption and emission bands and the relatively long luminescence lifetimes commonly observed in trivalent lanthanide ions are the principal advantages of these materials compared to other organic materials and quantum dots. Furthermore, the specific properties of the various host materials make it possible to configure rare-earth doped systems providing a wide variety of additional advantages. For instance, we highlight the high resistance to blinking or photobleaching of some materials, allowing wide applications in forensic sciences, biomedicine (bioimaging, photo-dynamic therapy or laser heating), optical temperature and/or pressure sensors [15,16,17,18,19,20].

Oxides with general formula La₃BO₇ (B = Os, Re, Nb, Ta and other transition metals) manifest some interesting catalytic, dielectric and magnetic properties [21,22,23]. Due to their comparatively high melting points, fine particle size, chemical stability and proton conductivity at high temperatures, La₃TaO₇ and La₃NbO₇ are proposed as materials for thermal barrier coatings [24] or as electrolytes [25,26] in some high temperature electrolysers. The weberite like structure of La₃NbO₇ compounds is an anion-deficient fluorite-related superstructure with Cmcm space group [27]. The structure is usually described as formed by slabs of c-axis oriented chains of tilted NbO₆ octahedra, sharing corners with parallel rows of lanthanum dodecahedra (one third of La ions in 8-fold oxygen coordination) edge-sharing parallel to (100). The remaining La ions, having 7-fold oxygen coordination, lie in between these layers. La ions can be substituted by other rare earth (RE) ions of similar radius.

In a previous work [28], we report the optical properties of the Er³⁺-doped La₃NbO₇ ceramic compound, checking the properties of this system from a structural point of view; we consider the viability of incorporating luminescent ions in this matrix and the effect of the impurities on the structure, etc., through the erbium luminescence. In this work, we explore the capabilities of the upconverted green emission of phosphors La_3-NbO₇:Er³⁺ as an optical temperature sensor working in the VIS range, under NIR excitation for two different temperature ranges: those with applications in biological processes (303–318 K) and biomedicine and in other higher temperature ranges (up to 525 K) relevant for industrial applications.

2. Materials and Methods

Following the synthesis route described in Reference [29], we prepared powder samples with nominal composition La_3−xEr_xNbO₇ (x = 0.01, 0.05 and 0.1) by the solid state reaction method. Stochiometric amounts of La₂O₃ (99.99%), Er₂O₃ (99.9%) and Nb₂O₅ (99.9%) Sigma-Aldrich (Darmstadt, Germany) powders were used as precursor. Rietveld’s analysis was carried out for X-ray diffraction patterns obtained from the samples confirmed by the orthorhombic Cmcm space group where the erbium ions were randomly allocated in the lanthanum sites 4c and 8g, surrounded by 8 and 7 oxygens, respectively. The X-ray diffraction patterns of a series of sintered samples with increasing amounts of Er₂O₃ powders previously reported in [28] prove the formation of the La_3−xEr_xNbO₇ phase (x = 0.01, 0.05, 0.1), with the largest peaks indexed with the type structure weberite based on the Cmcm space group. However, (see Supplementary Materials), some additional peaks appear in the XRD patterns and the foreign reflections correspond to a small amount of LaNbO₄ secondary phase. The relative percentage of this secondary phase increases with the amount of erbium in the samples. This fact appears to be inherent to the fabrication route in which the presence of Er³⁺ disturbs the necessary secondary La₂O₃ + LaNbO₄ → La₃NbO₇ reaction and eventually becomes responsible for the segregated phases [28,29]. Samples of the nominal formula La_3−xEr_xNbO₇ (x = 0.1), having 9.6% of LaNbO₄ secondary phase, were used in this work in an effort to maximize the emission intensity. Although present in our samples, the mentioned minority phases do not play a relevant role in our study. Additional details of sample preparation and structural analysis were described in Reference [28], together with the room temperature study of the luminescence of the Er³⁺ ion in this host matrix.

Up-conversion luminescence experiments presented in this work were performed in the temperature range from RT to 525 K in a tube-shaped electric furnace with the sample placed in its centre. The heating rate was 1 K/min to maintain thermal stability in the sample. The furnace temperature was monitored with a K-type thermocouple positioned close to the sample and connected to a voltmeter Fluke Calibrator 714 (Fluke Calibration, Everett, WA, USA). Erbium up-converted luminescence was excited using a tunable Ti-sapphire cw laser Spectra Physics 3900S (Fluke Calibration, Everett, WA, USA) pumped by a 532 nm diode laser (Spectra Physics Millennia Prime Laser 15SJSPG). The excitation wavelength was selected to be 975 nm. To avoid overheating the sample, the laser power was kept as low as possible. Through a lens system at the furnace exit, the up-converted green emissions were collimated and then focused on an optical fiber coupled to a 0.3 m focal length single grating spectrometer (Andor Shamrock 303i). The light was detected by a Peltier cooled silicon CCD camera (Andor Newton DU490A-1.7) with a resolution of about 0.7 nm (about 25 cm⁻¹ in the green region) and an integration time of 10 s. All spectra were corrected for the spectral response of the equipment. Conventional room temperature luminescence measurements were performed using a 488–490 nm laser diode as an excitation source and emission signal was detected with the experimental set-up described elsewhere [28].

Fluorescence Intensity Ratio (FIR) Technique

FIR technique uses the ratio between the intensities of the emissions coming from two TCL to characterize a temperature sensor in the following way [9]. Consider the luminescence of two TCL, having E₁ and E₂ energies, to a final level with energy E₀ (see Figure 1) being I₂₀ and I₁₀ their respective integrated intensities. By describing the population of both levels in thermal equilibrium using Boltzmann’s distribution, we can define a thermometric parameter ∆ as [9]:

$$\Delta =\frac{{\mathrm{I}}_{20}}{{\mathrm{I}}_{10}}=\frac{{\omega }_{20}^{\mathrm{R}}{\mathrm{g}}_2{\mathrm{E}}_2}{{\omega }_{10}^{\mathrm{R}}{\mathrm{g}}_1{\mathrm{E}}_1}{\mathrm{e}}^{{-\mathrm{E}}_{21}/\mathrm{KT}}{=\mathrm{C}\mathrm{e}}^{{-\mathrm{E}}_{21}/\mathrm{KT}}$$

(1)

where K is the Boltzmann constant, E₂₁ = E₂ ‒ E₁ is the energy separation between the thermalized emitter levels, g₂ and g₁ their respective degenerations (2J + 1) and ${\omega }_{\mathrm{ij}}^{\mathrm{R}}$ is the spontaneous emission rates that can be calculated using the theory of Judd–Ofelt [30,31,32,33,34]. Besides accuracy, two relevant figure quantities characterize a thermal sensor: the relative sensitivity and the resolution. The sensor relative thermal sensitivity S_REL (in % K⁻¹) is defined using the following expression [35]:

$${\mathrm{S}}_{\mathrm{REL}}=\frac{1}{\Delta }\left|\frac{\mathrm{d}\Delta }{\mathrm{dT}}\right|\times 100=\left(\frac{{\mathrm{E}}_{21}}{{\mathrm{KT}}^2}\right)\times 100$$

(2)

This parameter has the advantage of allowing the comparison of the performance with temperature of different optical temperature sensors regardless of their nature. From Equation (2), larger energy gap between the TCL levels implies higher thermal sensitivity. Nonetheless, this is true only to a certain extent, because larger energy gap also means lower population in the upper level, leading to lower intensity emission coming from such level [9,35].

Concerning the temperature resolution, it is usual to define the temperature uncertainty, δT (in K), as the smallest temperature change that can be achieved in each measurement. The temperature uncertainty can be estimated according to the equation: [35]:

$$\mathsf{\delta }\mathrm{T}=\frac{1}{{\mathrm{S}}_{\mathrm{REL}}}\frac{\mathsf{\delta }\Delta }{\Delta }$$

(3)

where S_REL is the relative sensitivity and the δ∆/∆ is the relative uncertainty on ∆, which for typical sensitive detectors (CCDs, among others) is around 0.03% [35].

3. Results and Discussion

3.1. Luminescence

VIS-NIR conventional luminescence of Er³⁺ ions obtained from a La_2.9Er_0.1NbO₇ powder sample excited at 488 nm and the upconverted luminescence obtained under the 975 nm excitation, both recorded in ambient conditions, are shown in Figure 2. A simplified level scheme is included on the side of Figure 2 showing the main excitation channels in which these VIS-NIR emissions are originated. In the VIS region, we observe the characteristic green bands corresponding to the (²H_11/2,⁴S_3/2) → ⁴I_15/2, transitions, dominating the spectra, together with a less intense red emission band centred at 660 nm corresponding to the ⁴F_9/2→⁴I_15/2 erbium transition. A low NIR luminescence band centred at about 850 nm, lying on the first biological window and corresponding to the ⁴S_3/2 → ⁴I_13/2 transition, is also obtained. From the analysis of integrated intensity of upconverted emission versus excitation power, a slope of 1.5 is obtained (see inset in the lower side), indicating that at least two IR photons are involved in the process in accordance with previous data [28,36,37]. We must note that the same bands are obtained in both spectra, but they have different relative intensities, depending on the luminescence origin. Obviously, the spectrum at the bottom of this figure was not obtained with low-power laser excitation. Thus, the effective temperature of the sample is higher in this spectrum than in that of conventional luminescence (upper side of the figure), originating from the 530 nm band having higher relative intensity. In all luminescence channels, the luminescence consists of broad bands, formed by a strong spread of narrow lines that extend to energies lower than usual values in other oxides. This fact is indicative of the Er³⁺ incorporation in a system with occupation of several available centres. Indeed, some intense narrow emission peaks observed in the low energy side of each emission band, those centred at 573, 695 and 883 nm, appear as characteristic features of the Er³⁺ luminescence in these samples. As the La₃NbO₇ matrix has two La sites with different oxygen coordination, we can assume that this widespread characteristic luminescence is mainly due to the emission of erbium ions incorporated in both sites, as was previously indicated [28]. However, because samples are fine powders in which the discontinuity in grain boundaries supports several types of defects, and due to the presence of LaNbO₄ secondary phase in the samples used in this work, the origin of the emission peaks is not definitively clear. Finally, we must emphasize that the presence of the secondary LaNbO₄ phase does not seem to play a significant role in this study, since we have not found differences in the luminescence behaviour in samples with different erbium concentration La_3−xNbEr_xO₇ with x = 0.01, 0.05 or 0.1) nor in samples with the same erbium nominal concentration (x = 0.1) synthesized in different series; so, the measurements’ repeatability appears to be guaranteed.

3.2. Optical Sensor Calibration

We have studied the temperature dependence of all upconverted emission bands. Concerning the ~850 nm NIR luminescence lying on the first biological window, our primary interest was focused on the physiological range of temperature (303–318 K). Unfortunately, we found that both intensity and shape of the 850 nm band are mainly insensitive to the temperature changes in this narrow range of temperature.

Since the green upconverted emissions of La_3−xNbO₇: Er³⁺ (x = 0.1) phosphor present high intensities and quantum yields along with the hypersensitivity of the upper thermalized energy level ²H_11/2 [30,38], we analysed the feasibility of using this powdered material as a temperature sensor, working not only in the VIS range belonging within the physiological range of temperature but also at high temperatures.

In this regard, the temperature evolution of the green upconverted emissions related to the (²H_11/2,⁴S_3/2) → ⁴I_15/2 transitions of Er³⁺ ions were measured in this sample, and the obtained spectra are shown in Figure 3. In these spectra it can be observed that as the temperature increases, the intensity of the emission band related to the transition ²H_11/2→ ⁴I_15/2 also increases. In addition, the ⁴S_3/2 → ⁴I_15/2 emission band decreases due to the thermally induced population from this level to the upper one. In this sense, the relationship between the intensity areas of these green upconverted emission bands related to the (²H_11/2,⁴S_3/2) → ⁴I_15/2 can be considered as I₂₀ and I₁₀ in order to obtain the “upconverted” radiometric thermometric ∆ in Equation (2).

The temperature evolution of ∆ in the whole range of temperature is shown in Figure 4. It is clearly observed that such parameter increases as the temperature does, describing a curve that can be simulated using the Boltzmann’s distribution law (Equation (1)). Using a standard least-squares fitting procedure, the values of C = 5.96 and an energy gap E₂₁ = 768 cm⁻¹ were determined. This data and the high thermal stability of this compound make it to be a good candidate for optical measurements of temperature in a range wider than that studied here.

Let us now focus our attention on the low temperature side of the range studied in this work. The relative thermal sensitivity S_REL as a function of temperature T, as defined by Equation (2) and also shown in Figure 4 in the biological range of temperatures, shows relatively high values of around 1.26 and 1.19 % K⁻¹ at 295 K and 303 K, respectively, with a low temperature uncertainties δT of around 0.05 K. The capabilities of the La_2.9Er_0.1NbO₇ phosphor as an optical temperature sensor are compared to other Er³⁺ and Er³⁺-Yb³⁺-based optical temperature sensors listed in

Table 1. Relative S_REL thermal sensitivities, temperature range and excitation wavelength of different nanomaterials doped with Er³⁺ ions, or co-doped with Yb³⁺ ions, used in optical temperature sensing in the biological temperature range. Classification has been done according to the value of the relative sensitivity S_R calculated at 303 K.

Er3+ Doped and Er3+-Yb3+ Based Host Matrix	λexc (nm)	T Range (K)	FIR Equation	SREL (%K−1) at 303 K	Ref.
BZT-BCT: Er3+ ceramic	980	200–443	9.97•exp(−1135/T)	1.24	[44]
Y2O3: Er3+-Yb3+ nanoparticles	980	273–480	13.30•exp(−1131/T)	1.23	[45]
ZnMoO4:Yb3+-Er3+-Li+	980	300–400	14.27•exp(−1111/T)	1.21	[46]
NaYbF4: Er3+	980	293–593	10.78•exp(−1098/T)	1.20	[43]
La3-xNbO7: xEr3+ (x = 0.1) phosphor	975	295–525	5.96•exp(−1097/T)	1.19	This work
KLa(MoO4)2: Yb3+-Er3+ phosphor	980	303–423	21.95•exp(−1079/T)	1.18	[42]
YAP: Er3+-Yb3 nanoperovskites	980	296–600	5.77•exp(−1073/T)	1.17	[13]
Y3Ga5O12: Er3+ nanoparticles	488/800	300–335	11.04•exp(−1047/T)	1.16	[47]
NaYF4: Er3+-Yb3+ nanocrystals	980	303–483	30.80•exp(−1034/T)	1.13	[48]
BaMoO4: Er3+-Yb3+ phosphors	980	300–575	20.06•exp(−980/T)	1.07	[41]
Tellurite glass: Er3+-Yb3 (7.5 mol% TiO2)	785	318–431	8.10•exp(−962/T)	1.05	[49]
CaWO4: Er3+-Yb3+ phosphors	980	303–753	17.20•exp(−944/T)	1.02	[39]
Chalcogenide glass: Er3+-Yb3	1060	296–438	8.85•exp(−962/T)	1.01	[7]
Na0.5Bi0.5TiO3: Er3+-Yb3+ ceramic	980	93–613	4.71•exp(−827/T)	0.90	[50]
Na2Y2B2O7: Er3+-Yb3+ phosphor	980	300–613	27.60•exp(−766/T)	0.83	[40]
NaLuF4:Yb3+-Er3+-Ho3+	975	298–503	1.46•exp(−654/T)	0.71	[14]
Silicate glass: Er3+-Yb3+	980	296–723	3.65•exp(−593/T)	0.64	[51]

. From this table, it is clear that this material presents good performance with temperature showing high thermal sensitivities compared to other phosphors, for instance CaWO₄:Er³⁺-Yb³⁺, Na₂Y₂B₂O₇:Er³⁺-Yb³⁺, BaMoO₄:Er³⁺-Yb³⁺ or KLa(MoO₄)₂:Er³⁺-Yb³⁺ phosphors [39,40,41,42], among others, at high temperature range. On the other hand, La_3−xNbO₇Er_x (x = 0.1) phosphor also shows high thermal sensitivities within the physiological temperature range. In this narrow range (303–318 K), the sensitivity is comparable to other materials reported for applications in such range; for example, YAP: Er³⁺-Yb³⁺ nanoperovskites [13], which also exhibit applications in laser heating, or NaYbF₄:Er³⁺ powders [43], which can be used for biological purposes, such as photo-thermal therapy or even drug release.

The obtained SREL values suggest the application of this material as an optical temperature sensor working in the VIS range, and the obtained parameters suggest that this material is not only interesting from the point of view of industry applications at high temperatures but also within the physiological range of temperature for applications in biological systems.

4. Conclusions

The temperature dependence of green-upconverted emissions of Er³⁺ ions in La₃NbO₇ phosphors was studied within both the physiological temperature range from 303 K to 318 K and at higher temperatures, up to 525 K, under excitation at 975 nm. The thermal sensitivity obtained for the green thermally coupled emission bands ranged between 1.26% K⁻¹ at 295 K and 1.19% K⁻¹ at 303 K. Compared to other optical temperature sensors based on Er³⁺ and Yb³⁺-Er³⁺ excited in the NIR range and working in the VIS range, these values are high enough to support the viability of the Er³⁺:La₃NbO₇ phosphor as an optical temperature sensor. Based on the FIR technique for both temperature ranges under study, the physiological temperature range with applications in biological systems and higher temperature ranges could be interesting for some industrial applications.