Nonlinearity of the Upconversion Response of Er³⁺ in Y₂TiO₅:Er³⁺,Yb³⁺ Ceramics When Varying the Wavelength of Incident NIR Excitation Radiation

Abstract

The upconversion response of Er³⁺ sensitized by Yb³⁺ in various crystalline hosts and illuminated with a laser light at around 980 nm revealed certain spectral shapes that are typical for each of the crystalline matrices containing the dopants. The purpose of this work was to measure the upconversion response of Er³⁺ as a dopant in Y₂TiO₅, sensitized by Yb³⁺, at different concentrations relative to the substituted Y³⁺ ion, and to reveal the subtleties of the mechanisms of the energy transfers between them and the lattice. Therefore, we synthesized Y₂TiO₅ ceramic samples doped with different concentrations of Er³⁺ and Yb³⁺, below 10% (mol), in order to minimize the distortion of the lattice. The oxide powders, obtained using the sol–gel method, as well as the ceramics were structurally and morphologically characterized using an X-ray diffraction analysis and scanning electron microscopy. When the ceramic samples were irradiated with an NIR laser light, it was found that, at a wavelength variation of only 2 nm of the incident radiation, from 973.5 nm to 975.5 nm, the upconversion spectra differed significantly. This nonlinearity is notable because it is not present in the case of other crystalline host matrices studied by us since the literature lacks information on this subject. We also correlated this effect with the simulated distribution of the average distances between Er³⁺ and Yb³⁺ ions in the host matrix.

1. Introduction

Upconversion (UC) is the process by which a system, be it an atom, an ion, or a molecule, through absorbing low-energy quanta from the environment, either simultaneously or in successive steps reaches a state from which the de-excitation, either to the ground or to an intermediate level, is achieved by emitting a higher energy quanta compared with the absorbed ones. The energy absorption process is a succession of events like ground-state absorption (GSA), and one or multiple excited-state absorptions (ESAs). The absorptions of the energy quanta could be resonant, phononic-assisted, sensitized, or cooperative sensitized etc. [1].

Different nanoparticles with UC properties, designed for either high conversion efficiency or tunable color, have been studied for biomedical applications, such as biosensing and bioassays, high-contrast agents, drug delivery, and various therapies [2]. In this area, the use of IR radiation has some major advantages over its counterparts that use higher energy radiation (UV or blue light) due to a deeper penetration into biological tissue and because it is much safer.

The UC of the Er³⁺:Yb³⁺ pair, in various crystalline and biological compatible hosts, is used in medical engineering, where it acts as an in situ nanoscale, high-fidelity temperature sensor, by probing the ²H_11/2,⁴S_3/2→⁴I_15/2 fluorescence emissions intensity ratio, which depends on the energy level populations, which in turn are influenced by temperature [3,4,5,6,7,8,9]. Also, Er³⁺:Yb³⁺ upconverting nanoparticles, due to their low energy excitation wavelength, can act as 980 nm excited triggers for drug delivery when combined with drug carrier particles [10].

In optical engineering, e.g., in fiber optics signal amplifiers, Yb³⁺ is used to enhance the Er³⁺ emission of ⁴I_13/2→⁴I_15/2 (1500 nm), using 980 nm as the pumping radiation, which is cheaper to obtain and has better efficiency [11,12,13,14,15].

Er³⁺:Yb³⁺ UC nanoparticles were also applied in pollutant detection and removal [16], efficiency increasing of the photocatalytic activity of TiO₂ [17], lamp phosphors, display panels, catalysts, sensors, and photovoltaic process enhancers [18].

The main goal of the studies of the Er³⁺:Yb³⁺ activator–sensitizer pair is to increase the UC efficiency of Er³⁺ in incident near-infrared (NIR) radiation. The NIR light is easily obtained today with cheap laser diodes. Many crystals were used as hosts for this ionic pair, e.g., Y₂O₃, BaGd₂ZnO₅, and NaYF₄ [19,20,21,22,23], each having a certain efficiency that is linked to the phononic interactions between Er³⁺ and Yb³⁺ and the crystalline lattice host.

Nevertheless, a subtlety of the mechanism of the UC process, i.e., why the red–green emission ratio of Er³⁺ ions is higher when the relative concentration of Yb³⁺ to that of Er³⁺ is increased, is not well understood. Only one paper was found addressing this issue [24], but it is not explanatory enough.

In order to better grasp the causes of this phenomenon, another crystal host, namely Y₂TiO₅ ceramic doped with Er³⁺ and Yb³⁺, is studied in this work. This ceramic was chosen not only because, to the best of our knowledge, the literature lacks information regarding the UC of Er³⁺ embedded in this matrix, but also because Y₂TiO₅ is special as the titanates display a wealth of useful characteristics like a large bandgap (≥3.3 eV [25]), thermal stability [25], photocatalytic properties, and they do not pose biological hazards [26]. Also, yttrium titanate ceramics display various phase states, which are very useful in the fission energy industry [27]. They can accommodate large defects in the crystalline matrix induced by high-intensity particle radiation or by allowing small-molecule gases (helium, hydrogen) to penetrate the lattice [28].

We selected Er³⁺ as the activator and Yb³⁺ as the sensitizer because the ²F_5/2→²F_7/2 transition energy of Yb³⁺ is quasi-resonant with the ⁴I_11/2←⁴I_15/2, ⁴F_7/2←⁴I_11/2 or ⁴S_3/2←⁴I_13/2 transitions of Er³⁺ [29], ensuring an easier energy transfer between these ions [1]. Also, the Er³⁺ energy levels are appropriately distanced between themselves [29] in order for the excited states to have a sufficiently long lifetime, such that absorption of other quanta happens before de-excitation and the UC is facilitated.

The aim of this study was to investigate how the conformation of the Er³⁺, Yb³ doped Y₂TiO₅ crystalline matrix influences the UC spectral response of Er³⁺ upon illumination at 976 nm. Hence, we compared the upconversion spectra of Er³⁺ in Y₂TiO₅ (YTO) with those in other host crystalline matrices, either previously studied by us or reported in the literature.

During this endeavor, we found that the UC process for this Y₂TiO₅ matrix is not linear and displays high sensitivity to the wavelength of the incident IR laser light, with variations as small as 2 nm leading to notable changes in the spectral composition of the upconversion emission in the visible range. This behavior is peculiar for this crystalline lattice, since Er³⁺ present in other crystalline hosts (e.g., Y₂O₃, BaGd₂ZnO₅) displays more linear characteristics, as we observed in our research.

Herein, we report for the first time this nonlinear behavior of the UC process of Er³⁺:Yb³⁺ doped ceramics, nonlinearity that is linked to the crystalline structure of Y₂TiO₅ both at the nanoscale and to the shape of the surrounding coordination polyhedra at the locus of Y³⁺ substitution.

This research consists also in improving the understanding of the subtleties of the energy transfer between the dopant ions and with the lattice by correlating the experimental data with simulations of the dopant distributions in the crystal and their relative interionic distances.

This nonlinearity could find different applications, such as cheap laser wavelength indicators, checking of structural modifications at the nanoscale level in materials, detecting forgeries, 980 nm laser tuning, or doppler sensors for noninertial fiber gyroscopes.

2. Materials and Methods

2.1. Materials Used and Dopant Concentrations Labelling

The synthesis of Y₂TiO₅ ceramics, doped with Er and Yb samples, was carried out in two steps: first, by obtaining the precursor metallic oxides using the sol–gel technique; second, these oxides were then molded into pellets using high pressure, and the pellets were sintered at a high temperature and atmospheric pressure.

The reagents used, erbium(III) nitrate pentahydrate, ytterbium(III) nitrate pentahydrate, yttrium(III) nitrate hexahydrate, titanium(IV) butoxide 97% (TB), citric acid (CA), and ethylene glycol (EG) were purchased from Sigma Aldrich (Merck Group, Darmstadt, Germany) and were used as received.

Table 1. Percentages of Er³⁺ and Yb³⁺ dopant ions replacing Y³⁺ ions in Y₂TiO₅ and the label of the samples. The Er/Yb concentration ratios are chosen to be either powers of 2 or ratios of 2.

Dopant Concentration (%)		Sample Label
Er3+	Yb3+
0	0	YTO 0-0
0	1	YTO 0-1
1	0	YTO 1-0
1	2	YTO 1-2
1	4	YTO 1-4
1	8	YTO 1-8
2	4	YTO 2-4
4	4	YTO 4-4
3	6	YTO 3-6

presents the chosen concentrations of the dopant ions for replacing yttrium atoms in Y₂TiO₅ and the sample labeling that will be used throughout this text.

YTO 0-0 was chosen as the reference for the crystal structure when all the conditions of synthesis were the same. Also, YTO 0-0 was used to investigate the response of the pure Y₂TiO₅ matrix at a 976 nm illumination (fixing the background). YTO 0-1 was chosen to investigate the response of Yb³⁺ ions at the 976 nm range and observe the energy that the emitted photons have. The dopant concentration values were chosen to be small enough such that the Er³⁺ and Yb³⁺ ions should replace the Y³⁺ ions in the Y₂TiO₅ matrix with the lowest possible lattice distortions. As it will be seen, the lattice is very sensitive to these dopants, and concentration values higher than 10% (relative to Y³⁺) strongly hinder the equilibrium of the orthorhombic phase of Y₂TiO₅.

2.2. Synthesis of Oxide Powders

The metal nitrates in the appropriate quantities, CA (2 moles for every mol of metallic ions) and TB were dissolved in EG. The utmost care was taken for no traces of water to be in the reaction mixture since water induces the quick hydrolysis of TB. The reaction mixture was heated at 120 °C under continuous magnetic stirring (300 rpm) for slow evaporation until a transparent and thick gel was formed. The gel was thermally treated at 400 °C for 3 h until the most carbonaceous compounds were removed, and a grayish powder was obtained. The as-synthesized powders were calcinated at 900 °C for 3 h, after which their color changed from grey to pinkish due to the presence of Er³⁺ ions.

2.3. Obtaining of Er, Yb Doped Y₂TiO₅ Ceramics

Y₂TiO₅:Er,Yb ceramic pellets of 13 mm diameter and 1 mm thickness were obtained in two steps: first, the powders were pressed at 200 bar for 2 min, followed by grinding the resulting pellets in an agate mortar; second, the powders were pressed at 100 bar for 1 min with a slow release of the pressure. This pressing procedure was chosen to avoid the brittleness of the resulting ceramics because the initial powders had a very low cohesiveness. The first step ensures an aggregation of the particles into larger ones, whereas the second step eliminated the voids created between them, while inserting smaller inner tensions and preventing the pellet from cracking during manipulation and thermal treatment.

Two pellets for each composition were produced and placed, one on top of the other, in alumina crucibles. This stacking prevents the accidental modification (contamination, evaporation of constituents, etc.) of the composition of the touching faces of the pellets during the prolonged oven time.

The pellets were sintered at 1250 °C for 16 h. This is the minimum time for the sintering treatment for the Y₂TiO₅ crystal lattice stabilization. The temperature was chosen according to the reported results [30], which showed that the diffusion coefficients for oxygen ions in the Y₂TiO₅ structure are sufficiently high even at temperatures lower than the melting point. This ensures the formation of the orthorhombic phase while impending the segregation of phases with a lower energy of formation (−3.98 eV/atom for Y₂O₃ [25]) than Y₂TiO₅ (−3.87 eV/atom [25]) or other phases like hexagonal, pyrochlore, fluorite, or compounds like Y₂Ti₂O₇ and YTiO₃ [31].

2.4. Characterization of Y₂TiO₅ Samples

The oxide powders and the ceramics were characterized using X-ray diffraction (XRD) with a Rigaku Miniflex II, Tokyo, Japan, diffractometer with Cu Kα radiation in the 2θ range of 10°–70°, at a 2 °/min rate, and a step of 0.01°. The scanning electron microscopy (SEM) was performed using a Tescan Vega 3LM, Brno, Czech Republic, microscope equipped with an EDS spectrometer. The upconversion spectra were measured using a USB4000CG-UV-NIR spectrometer from Ocean Optics (Orlando, FL, USA) and the OceanView software version 1.6.7. The LCU98E042Ap diode (Laser Components GmbH, Olching, Germany) was driven with emitted power control and thermally stabilized because the lasing wavelength varies with temperature. The thermal control allowed for the IR laser wavelength tunability with subnanometer precision. The ray was collimated with a standard laser collimator, which illuminated an area of 0.5 mm², allowing for individual microcrystal probing. The acquisition fiber optic head entrance was at a distance of 3 mm from the illuminated spot and had a diameter of 200 µm.

The spectra acquisition periods for the photon counting were 30 ms, and the laser illumination power was 150 mW. All the measurements were averaged over 10 full-spectrum acquisitions, such that a single averaged spectrum was measured in 0.3 s.

These power and time parameters were chosen so as not to saturate the spectrometer sensor in the spectral range of interest (visible). The measurement setup for the upconversion spectra is depicted in Figure S1 (Supplementary Materials).

3. Results

3.1. Powder Oxides XRD Characterization

The XRD analysis of the resulting oxide powders reveals smooth and wide peaks, indicating the formation of a cubic phase, which is the precursor of the orthorhombic Y₂TiO₅ (formed after the sintering), and is presented in Figure 1.

3.2. XRD Analysis of Y₂TiO₅:Er,Yb Ceramics

The XRD patterns of the YTO 0-0, 0-1, 1-0, 1-2, 1-4, and 1-8 ceramics, as well as the simulated orthorhombic YTO 0-0 diffractogram using the VESTA 3 software [32], can be seen in Figure 2. The XRD patterns of the other Y₂TiO₅:Er,Yb samples are presented in Figure S2 (Supplementary Materials). It was found that the sintering process at 1250 °C for 16 h is the optimum thermal treatment to obtain the orthorhombic Y₂TiO₅ phase. Other attempts, i.e., a lower calcination temperature or shorter time, resulted in a mixture of cubic and orthorhombic phases, whereas higher temperatures gave rise to other phases segregation.

In the case of YTO 1-4, the matrix begins to be distorted (a small peak at 2θ = 30° can be observed), and for YTO 1-8, the distortions are even higher. These deformations are compatible with the undoped structure because of the large interstitial (no bond) spaces and the fact that Yb³⁺ ions have almost a 4% smaller ionic radii than those of Y³⁺ ions.

3.3. Crystallinity of the Ceramic Pellets

After sintering, the protected face of each ceramic pellet was first sanded with 2000 grit, then polished with 6000 grit until the face was gloss-shining. The procedure revealed, only under frontal illumination with a bright white light, the microcrystal structures of the pellets. An example of the polished face of YTO 1-0 is shown in Figure 3 (see also Figure S3), and one can see that the sizes of the microstructures are uniform.

The uniformity of the microcrystal size for each dopant concentration shows a successful sintering process. The variation in the average size of the microcrystals shows how the dopant concentration influences the crystallization process. There is a variation in the microstructure sizes across the cases of dopant concentrations, with the lowest feature sizes being for YTO 1-0 (Figure 3) and the largest ones for YTO 1-2 (Figure S3). This is an indication that the doping ions (even if the total dopant concentration relative to Y is below 10%) strongly influence the crystallization process and the crystalline phase formation (Table S1) revealed by the XRD patterns (Figure 2).

3.4. Structure of the Y₂TiO₅:Er,Yb Crystal Unit Cell

Y₂TiO₅ crystallizes in the orthorhombic system with P1 symmetry. The unit cell parameters are: a = 3.72 Å, b = 10.45 Å, and c = 11.35 Å. The CIF data [25] regarding the Y₂TiO₅ compound were visualized using the VESTA 3 software [32] and reveal the positions of the ions in the unit cell, as depicted in Figure 4.

Yttrium ions are coordinated by seven oxygen ions, and Ti⁴⁺ ions are coordinated by five oxygen ions. All Y³⁺ ions have the same geometry of the coordination polyhedra.

It can be observed (Figure 5 and Figure S4) that the orthorhombic matrix has large voids, which can easily give rise to lattice defects, nanostructural instability, the accommodation of other kinds of ions, or being permeable to gases with low atomic radii. The ionic radii for the involved metallic ions, coordinated by seven oxygen ions, are provided in

Table 2. Ionic and crystal radii for Y³⁺, Er³⁺, and Yb³.

Ion in VII Coordination	Crystal Radius (Å)	Ionic Radius (Å)
Y3+	1.100	0.960
Er3+	1.085	0.945
Yb3+	1.065	0.925

. Yb³⁺ has a 4% smaller radius compared with Y³⁺ [33]. Therefore, larger concentrations of Yb³⁺ hinder the shape and stability of the orthorhombic Y₂TiO₅.

The XRD phase matching analysis of the samples performed using the Match! software [34] shows that, for each ceramic, there is a combination of phases that are compatible with the orthorhombic structure of Y₂TiO₅ (Table S1). This is due to the interstitial spaces of the structure, as seen in Figure 4 and Figure 5, which allow for such a variability of combinations of phases with the same structure of the unit cell but with slightly different geometrical parameters; the percentage of the combinations is also influenced by the dopant concentrations. In the case of YTO 1-8 and YTO 3-6, which have high dopant concentrations, fluorite Y₂Ti₂O₇ and even hexagonal Y₂TiO₅ were formed.

Consulting the phase diagram for Y₂O₃–Y₂Ti₂O₇ [31], one can conclude that the phase composition is very sensitive to the types of doping ions and their concentrations. Even low concentrations of doping ions can significantly alter the phase equilibrium, and this is due to the metastable conformation of the matrix.

In the cases of YTO 1-8 and YTO 3-6 with a high concentration of Yb³⁺, the XRD analysis showed (Figure 2 and Figure S2), beside the expected orthorhombic phase, a hexagonal phase and the Y₂Ti₂O₇ fluorite phase. The YTO 1-8 ceramic was obtained also trying slightly different sintering conditions (at a maximum temperature of 1300 °C), but the hexagonal and fluorite phases appeared in all cases, which is an indication that Yb³⁺ promoted the formation of the Y₂Ti₂O₇ phase, supposedly because of its lower ionic radius (0.925 Å) than that of the Y³⁺ radius (0.960 Å).

3.5. SEM Investigation of the Sintered Powders and Ceramics

The SEM images show that the powders (calcined also at 1250 °C) and the ceramics have very similar aspects, with the pellets being more compact (Figure 6 and Figure S5). Comparing with other morphologies shows that the closest matching one is that of volcanic ash [35,36].

In all cases, there is no evidence of any forms of large-scale crystals, but only a mix of irregular shards with a large range of dimensions, which is an indication that the good crystal structure, revealed using XRD, is only at the nanoscale level and that these nanocrystals are attached among themselves in a macroscopic phase.

This aspect is allegedly due to the loose structure of the YTO crystal unit cell, which allows for deformations and other inter-nanocrystalline transition phases, which could also be related to the variate compositions revealed using the X-ray diffractograms (Table S1). Also, it can be observed that the higher the total dopant concentration, the smaller the ceramic grains, while the aspect of the powders remains the same (Figure S5).

3.6. Upconversion Light on a Wide Area

Prior to the upconversion emission spectroscopic investigation, the pellets were tested by illuminating them with a grazing IR laser light with two wavelengths, 973.5 nm and 975.5 nm, over an area of about 3 mm × 3.5 mm with a power of 175 mW. The pictures are presented in Figure 7 for YTO 1-2 (for the other cases, see Figure S6). The microcrystals shine the upconverted radiation differently, revealing a mosaic-like local variation in brightness (best seen in the YTO 1-2 case, Figure 7A,B) and bright spots in the other cases (Figure S6).

This brightness variability is due to the size of the crystals and not caused by dopant concentration nonuniformity, because the EDX analysis (Figure S7) revealed a homogeneous distribution of the dopants. This is notable, especially in the cases with higher concentrations of Er³⁺, YTO 2-4, YTO 4-4, and YTO 3-6, with spots having slightly green or red hues, yet the spectral shapes being the same. The stripe variation in brightness is due to the unevenness of the laser field emitted by the diode.

How the dimension of the microcrystal influences the upconversion mechanism remains to be established, but the images strongly suggest that the sizes of the microcrystals are influencing the phononic energy transfers and/or nanodot-like energy confinement.

Reorienting the pellets relative to the polarization of the laser light shows no variability in the brightness of the spots, underscoring the idea that the energy transfers are isotropic, not depending on some preferred light–crystal relative orientations.

3.7. Upconversion Spectra

The specifics of the UC spectra measurement we used (pinpoint focusing the laser light) allowed us to probe the individual spots on the pellets and check the variability of the response spectra across their surface.

The brightest spots on the surfaces of the pellets were targeted with 150 mW illumination power (continuous wave) and the spectra were measured for two close wavelengths, 973.5 nm and 975.5 nm. Figure 8 and Figure 9 present the upconversion spectra for the visible part for both values of the excitation radiation wavelength.

The difference between these excitation energy values is about 20 cm⁻¹, but the effects on the upconversion response are important. This fact is an indication that the upconversion mechanism can be linked to some detuning of some yet-unspecified resonant energy transfers between Er³⁺ and Yb³⁺ ions. Not only are the upconversion spectral compositions different for the two incident wavelengths, but also the emitted power curves are altered.

This high sensitivity with the incident radiation wavelength of Er,Yb:YTO ceramics is different from the other ceramics (Y₂O₃, BaY₂ZnO₅, BaGd₂ZnO₅, CaGd₂ZnO₅, BaY₂O₄, and BaY₂MgO_5, for the same dopant concentration) that we tested. In the 973.5–975.5 nm excitation wavelength range, these ceramics display a similar upconversion spectra with a linear dependence of the emitted power vs. incident power, unlike Y₂TiO₅.

Figure 10 presents the peaks (between 840 nm and 882 nm) of the emission of the transition ⁴S_3/2→⁴I_13/2 of Er³⁺ and the peaks between 900 nm and 940 nm, which are from the anti-Stokes sideband generated by the emission of Yb³⁺ when decaying from ²F_5/2 to ²F_7/2. Their widths and intensities are indications that the phononic interactions are significant. Also, they help to estimate the phonon energy for Y₂TiO₅ at about 610 cm⁻¹.

It can be seen that, in the case when the excitation was performed at 973.5 nm radiation, the anti-Stokes peaks for Yb³⁺ are almost double those in the case with 975.5 nm. The YTO 1-8 case is an exception, revealing a diminished interaction between the phonons and Yb³⁺ ions.

One can observe in Figure 10 the variability in the ⁴S_3/2→⁴I_13/2 transition of Er³⁺ across the dopant concentrations and also in the wavelength of the exciting radiation. For the 975.5 nm excitation, it is observed that a higher Yb³⁺ concentration promotes the decay to the ground of Er³⁺, instead of being trapped to ⁴I_13/2. At the 973.5 nm excitation, it is seen that Er³⁺ has a greater tendency to reach ⁴I_13/2, only 8% of Yb³⁺ strongly hindering that. The mechanism by which the Yb³⁺ ions are influencing the Er³⁺ energy level transitions, when embedded in Y₂TiO₅ crystal matrix, will be the subject of further research.

4. Discussion

4.1. Fitting the Upconversion Spectra

The spectra with the most intense peaks of Y₂TiO₅:Er,Yb ceramics were chosen for fitting. These were YTO 1-4 for the green part (⁴S_3/2→⁴I_15/2) and YTO 1-8 for the red part (⁴F_9/2→⁴I_15/2). The fitting squared error is under 10⁻⁹, and the results are shown in Figure 11. The fitted peaks data are presented in the Supplementary Material (Tables S2 and S3). The homogeneous broadened Lorentzian profiles and the relatively large FWHM are indications that the de-excitation of the Er³⁺ ions is influenced by phononic interactions. While the two Stark manifolds of ⁴S_3/2 are easily detectable (separation at 18000 cm⁻¹), the five ones in the case of ⁴F_9/2 are a bit harder to separate.

4.2. Comments on the Upconversion Spectra

When only Er³⁺ is present, the UC is very weak, as can be seen in Figure 8 and Figure 9. The presence of Yb³⁺ ions greatly increases the efficiency due to the enhanced capture of the incident NIR radiation. The fitted peaks are Lorentzian, and the broadening of the lines is homogeneous, which indicates that the Er³⁺ ions are trapped in a crystalline phase. The high value of FWHM shows that the local field strength is lower than in the case of Y₂O₃, indicating that the Er³⁺ ions are being more loosely trapped when replacing Y³⁺ in the Y₂TiO₅ lattice.

Using the VESTA 3 software [32], the average Er³⁺↔O^2- distances, i.e., the average radii of coordination, were calculated, and they are 2.305 Å in the case of cubic Y₂O₃ (C_2v site) (VI coordination) and 2.363 Å in the case of orthorhombic Y₂TiO₅ (VII coordination). If this small difference (2.5%) is to account for the large differences between the average widths of the spectral lines in the case of Y₂O₃ (average FWHM ≈ 33 cm⁻¹) [37] compared with those of Y₂TiO₅ (average FWHM ≈ 50 cm⁻¹), it remains to be investigated. Also, this suggests that the smaller lifetimes of the excited levels in the case of the Er³⁺ ions trapped in Y₂TiO₅ may have other causes than the immediate sphere of coordination.

4.3. YTO 1-2 Special Case

It is worth noting that the YTO 1-2 ceramic has peculiar behavior. The upconversion spectrum for the excitation at 975.5 nm has an almost identical shape to the Y₂O₃ doped with 1% Er³⁺ and 2% Yb³⁺ (Y₂O₃ 1-2) [37]. The position and intensity of the peaks are the same for the red emission, while for the green emission, minor differences are observed (Figure 12).

While the peak positions are the same, their intensity distributions show that, in the case of the 973.5 nm excitation, the upper level of the Stark-degenerated ⁴S_3/2 is more populated. This is unexpected because the Stark splitting, caused by the intensity and shape of the embedding crystal field, should be different in the two cases (Y₂TiO₅ vs. Y₂O₃), with the coordination polyhedra of Er³⁺ and the Er³⁺–O^2- distances not being the same.

Moreover, while the upconversion spectrum of Y₂O₃ 1-2 has the same shape for both the incident IR wavelengths (Figure 12), for YTO 1-2, the spectra differ drastically from 973.5 nm to 975.5 nm, as can be seen in Figure 8 and Figure 9. As the energy difference is only 20 cm⁻¹ between 973.5 nm and 975.5 nm, such behavior can only be due to other causes than the coordination polyhedra shape and dimension.

The only parameter that is common for both cases is the interionic Er³⁺↔Yb³⁺ average distance distributions, and the almost identical shape of the spectra in the 975.5 nm case is a strong indication that the upconversion mechanism has, at its roots, not only the general considered energy transfers between the Er³⁺ and Yb³⁺ ions, but also some interactions between the Er³⁺ ions and an overall photon field maintained by the Yb³⁺ ions, disregarding the peculiarities of the host matrix.

To the best of our knowledge, this behavior is not explained in the literature, its origin is unknown, and a segregation of some of the Y₂O₃ phase was suspected. The synthesis and analyses of YTO 1-2 were repeated three times, but each time the same results were found, with no Y₂O₃ phase shown in the XRD patterns, and the upconversion spectra being reproducible. The best indication that the Y₂O₃ phase is not segregated and cannot be the cause for the unusual spectrum is that, for the 973.5 nm case, the upconversion spectrum differs from that of the 975.5 nm case, as seen in Figure 8 and Figure 9 for YTO 1-2.

4.4. Comparison of the Upconversion Efficiency

The integral intensities for the entire visible emission (in the red and green domains), for both the illumination wavelengths, are presented in Figure 13. The graphs are for an illumination with 150 mW for 30 ms. One can notice that YTO 4-4 has the lowest efficiency in all cases, i.e., Er³⁺ quenches itself. YTO 1-4, which is the brightest at 973.5 nm, is less responsive to 975.5 nm. In the 973.5 nm case, the green emission is brighter and quickly becomes hindered at 975.5 nm. For an energy difference of only 20 cm⁻¹ between 973.5 and 975.5 nm, this is unexpected.

In Figure 14, the samples are ordered by decreasing the percentage of green in the total emission, clearly showing the influence of Yb³⁺ in shifting the population of the excited Er³⁺ ions toward ⁴F_9/2. It is remarkable is that the excitation at 973.5 nm induces brighter green than red emissions. However, the upconversion efficiency for the Y₂TiO₅ matrix is lower than that of Y₂O₃ [37].

From Figure 13 and Figure 14, one can observe that increasing the Yb³⁺ concentration (YTO 1-2, 1-4, 1-8) facilitates the transition from ⁴F_9/2→⁴I_15/2, whereas maintaining the Yb³⁺ content and increasing the Er³⁺ concentration (YTO 1-4, 2-4, 4-4) maintains the red intensity while the total emission is quenched because of the decreasing of the Er³⁺↔Er³⁺ distances.

4.5. Correlation with the Er³⁺↔Er³⁺ and Er³⁺↔Yb³⁺ Interionic Distances

Such a behavior must be correlated with the distances between the dopants, either Er³⁺ ↔Er³⁺ or Er³⁺↔Yb³⁺. This was performed using a numerical simulation of randomly placed Er³⁺ and Yb³⁺ dopant ions (substituting Y³⁺ ions) in a cubic lattice (500 Å edge) of crystalline orthorhombic Y₂TiO₅ and computing the distributions of the average distances from the N = 4, 6, 8, closest Yb³⁺ neighbors of each Er³⁺ ion to itself, as exemplified in Figure 15A. The examples for the Er³⁺↔Er³⁺ distance distributions in the case of 4% Er³⁺ are shown in Figure 15B and

Table 3. Peaks of the average distance distributions for the different percentile Er³⁺ and Yb³⁺ concentrations from a random Er³⁺ ion to the first group of 4, 6, and 8 neighbors, be they Er³⁺ (for 1-0, 2-0, 4-0) or Yb³⁺, as specified in the first column. The last line in the table shows the peak values for YTO 4-0, whose colored distributions and peaks are shown in Figure 15B (green; blue and magenta lines).

Ceramic %Er-%Yb	N = 4 Neighbors Peak at (Å)	N = 6 Neighbors Peak at (Å)	N = 8 Neighbors Peak at (Å)
YTO 1-8	6.5	7.5	8.3
YTO 3-6	7.3	8.0	9.0
YTO 4-4	8.5	9.5	10.5
YTO 2-4	8.5	9.5	10.5
YTO 1-4	8.5	9.5	11.0
YTO 1-2	10.5	12.0	13.0
YTO 1-0	13.0	14.8	16.0
YTO 2-0	10.5	12.0	13.0
YTO 4-0	8.5	9.5	10.3

.

Figure 16 presents the samples, ordered by the decrease of the average Er³⁺↔Yb³⁺ distances, and it can be seen that, in the case of YTO 1-2, 2-4, and 4-4, the average distances from the activators to the sensitizers are almost the same. Nevertheless, the variability of the spectral shape and compositions for the UC responses in the respective cases is a clear indication that the Er³⁺↔Er³⁺ interaction must be taken into consideration when designing for UC efficiency, because increasing the Er³⁺ content does not necessarily lead to better performance.

Figure 17 shows the total intensities for the two cases (973.5 nm and 975.5 nm) with the samples ordered by the decrease in the average radii (any of R₄, R6, and R₈), according to the graph in Figure 16. The drop in intensities for the YTO 4-4 case should be noted, clearly showing Er³⁺↔Er³⁺ quenching.

4.6. Emitted Power Variation Curves

The power of the incident IR laser radiation was increased from 25 mW to 150 mW, by steps of 25 mW. The same spot was illuminated and the UC spectra were measured for both wavelengths, 973.5 nm and 975.5 nm. The photoluminescent integral intensities, both for the green (⁴H_11/2,⁴S_3/2→⁴I_15/2) and red (⁴F_9/2→⁴I_15/2) emissions, are displayed in Figure 18.

It is interesting to observe that, in the case of 975.5 nm, the output power starts to saturate beyond 100 mW, while in the case of 973.5 nm, the linearity is preserved. This shows that, at a certain threshold of local power density for 975.5 nm, a certain percentage of Er³⁺ ions involved in the UC process either prefer other decay channels than directly to the ground state, or the emitted photons are absorbed (notable for the YTO 1-8 and YTO 3–6 samples).

Also, there is a strong variability in the concentrations of the doping ions for the output power for the two incident wavelengths. Notable is the case of YTO 1-4, where the output power for green emission at 973.5 nm is much higher than that of red emission (bright red (round dots) line in Figure 18A,B), while for YTO 1-8, the green emission is strongly suppressed (blue line (up triangles) in Figure 18C,D).

5. Conclusions

Erbium and ytterbium-doped diyttrium titanate ceramic samples were obtained at 1250 °C from oxide powders, which were synthesized using the sol–gel method. The characterization using XRD and SEM showed that the doped Y₂TiO₅ ceramics’ microstructure and phase composition were very sensitive to the dopant concentrations. When the Yb³⁺ concentration is increased, additional hexagonal and fluorite phases appear, which fixes the maximum threshold for the Yb³⁺ concentration at 10%, beyond which the Y₂TiO₅ orthorhombic phase becomes heavily distorted.

The samples were illuminated with 973.5 and 975.5 nm from a thermally tuned laser diode, and the resulting UC visible spectra were measured and characterized. In the case of Y₂TiO₅:1%Er³⁺:2%Yb³⁺, we observed that the upper Stark level of ⁴S_3/2 of Er³⁺ becomes more populated when illuminated with 973.5 vs. 975.5 nm, this fact inducing a notable change in the hue and peak intensities in the UC response. Also, the emitted green–red intensities, both relative and absolute, for the other samples were observed and compared for the 973.5 and 975.5 nm irradiations, showing notable differences that were not observed in other crystalline lattices that we studied or found in the literature. The curves relating the emitted intensity versus the excitation power of the incident radiation were not linear, with a saturation tendency beyond 100 mW for green (515–575 nm), in the case of 973.5 nm irradiation, and for red (640–700 nm), in the case of 975.5 nm illumination.

Simulations were performed for the distributions of the interionic distances for Er³⁺↔Er³⁺ and Er³⁺↔Yb³⁺, and the samples were ordered by the distance of the maximum for each case of the dopant concentrations. It is seen that Er³⁺↔Er³⁺ quenching takes place when Er³⁺ is 4%, regardless of the Yb³⁺ concentration, showing that Er³⁺↔Er³⁺ energy transfers are more prevalent than those between Er³⁺ and Yb³⁺. The fact that the upper Stark level of ⁴S_3/2 is more populated when the illumination is performed with 973.5 nm vs. 975.5 nm is puzzling, since the distance between the two levels of ⁴S_3/2, split by the crystal field of Y₂TiO₅, is more than 20 cm⁻¹, which is the energy difference between 973.5 nm and 975.5 nm.

The source of this energy difference will be the subject of further investigations; nevertheless, the sensitivity of the effect calls for practical engineering applications with a high degree of precision. Some of these are laser diode wavelength tuning, cheap laser diode quality control, laser wavelength indicators, crystalline structural modification indicators when YTO is incorporated in other bulk ceramics, cheap doppler sensors with under 2 nm precision detection for non-inertial optical fiber gyroscopes, and last but not least, forgery prevention by including certain YTO nanodots in the protected samples.