Impact of Sc3+-Modified Local Site Symmetries on Er3+ Ion Upconversion Luminescence in Y2O3 Nanoparticles.

Rare earth (RE) doped yttria sesquioxide has been widely used as host materials for upconversion (UC) phosphors due to their high refractive index, wide band gap, and high melting point. Meanwhile, while fluoride matrices with low phonon cutoff energies exhibit stronger UC emissions, RE-doped oxides exhibit better thermal stability and higher thermal sensitivity when applied as optical temperature sensors. In this work, Sc3+ is substituted in RE-doped Y2O3 lattices to generate smaller cation sites, enhancing the crystal field and modifying the allowed optical transitions. Er3+ is used as a photoluminescent probe to study the effect of site position and symmetry on the UC performance. In comparison with the traditional hydrothermal method, Sc3+ is successfully incorporated into the Y2O3 lattice via the co-precipitation/molten salt method without segregating observed. The Judd-Ofelt analysis was applied to determine the local symmetry and efficiency changes. Sc was found to be able to improve the luminescence performances of Er in Y2-x Sc x O3 (YScO) hosts by adjusting the local symmetry level around the luminescent sites. The local symmetry level was reduced with less than 30 mol % of Sc doping concentration based on the changes in Ω2 values. Meanwhile, the YScO oxide was found to significantly improve the luminescence intensity and red-to-green ratio at a lower Yb3+ concentration (5 mol %) instead of a higher concentration (20 mol %) commonly used. This was attributed to an increased energy transfer between the closer Yb3+-Er3+ pairs. Overall, this work allows the spatial occupancy of luminescence centers in the metal oxide host materials to optimize the UC luminescence performance and develop a high-efficiency oxide material for high-temperature applications such as optical thermometry.

Introduction

In recent years, upconversion (UC) luminescence of rare earth (RE)-doped nanophosphors has drawn considerable interest due to their potential applications in lasers,[1,2] solar cells,[3−5] biological fluorescence imaging and detection,[6,7] IR quantum counters,[8] and display technologies.[9] Among the luminescent RE elements, Er3+, Ho3+, and Tm3+ are commonly used as UC activators due to their ladder-like arranged energy levels, long lived excited states, and excellent resonance.[10,11] Meanwhile, Er3+ shows particularly high UC efficiency owing to its similar sequential energy gaps, e.g., 4I11/2 and 4I15/2 states and 4F7/2 and 4I11/2 states, for sequential IR excitation,[10,11] which can be enhanced through the incorporation of sensitizers (i.e., Yb3+), reducing the possibility of non-radiative relaxation. Yb3+ sensitization is commonly used due to its large absorption cross section (9 × 10–21 cm–2, 2F7/2 → 2F5/2) and simple energy-level structure with only one excited state of 2F5/2, which matches well with those of many RE ions.[12] Additionally, a number of common dopants have been investigated to improve the luminescence properties, including Li+, Sr.+, Ca2+, Sr2+, etc.[13−16] However, care must be taken to avoid changing the crystal structure (Sr/Ca) or prevent unwanted energy transfer, which leads to enhanced quenching (Li).[17]

Alternatively, the UC efficiency can be improved by modifying the host material to prevent non-radiative relaxation via parasitic surface sites or phonon vibrations.[10] Fluorides are widely applied as host matrices for UC phosphors, but their low thermal stability limits their application.[11] Thus, the challenge is to design a material with high chemical and thermal stability and strong UC emissions comparable to fluoride phosphors. Here, pure RE oxide hosts are applied as a host material for UC phosphors due to their high melting points, high refractive indexes, large band gaps, excellent physical and chemical stability, and low phonon energy.[18−23] Specifically, yttrium oxide (Y2O3) forms a sesquioxide with C2 (octahedral) and C3 (trigonal prismatic) point symmetries.[24] One unit cell contains 32 cation sites, with 24 of these with C2 point symmetry and 8 with C3.[24] For luminescence applications, the optical emissions are assigned to the electronic transition of the dopant ions in the C2 sites, while those of the dopants in the C3 sites are forbidden due to the inversion symmetry. Similar to Y2O3, Sc2O3 has the same cubic bixbyite structure with a smaller cation size (Y3+-102 pm vs Sc3+-74.5 pm), creating smaller dopant sites and an enhanced local crystal field. Meanwhile, it is known that the luminescence performance can be improved if the local structure of the RE site has a lower symmetry level.[25] As such, the larger RE ions occupying the C2 sites are expected to prompt higher luminescence intensity of dopant ions. Co-doping the Y2O3 system with Sc3+ allows for manipulating lattice site symmetry to concentrate Er3+ in the lower symmetric C2 sites.

In this work, Y2–ScO3 (YScO):Er3+ (5 mol %) and YScO:Er3+ (5 mol %),Yb3+ (y mol%, y = 5, 10, 15, 20) nanoparticles (NPs) were synthesized via a co-precipitation/molten salt synthesis process with varying Sc3+ molar ratios. Structural characterization was performed using XRD to affirm the doping homogeneity and crystallinity of the NPs. Next, the optical performance was determined using photoluminescence (PL) spectroscopy, showing that the YScO:Er3+ NPs with low Sc concentration (x < 0.5) exhibit improved UC intensity. To determine the Sc effect on the UC efficiency, Judd–Ofelt calculations were applied to each same using the PL excitation spectra measured at 77 K.[26]

Experimental Section

Sample Preparation of Yttrium Scandium Oxide (YScO) NPs

YScO:Er3+ and YScO:Er3+,Yb3+ NPs were synthesized using both the co-precipitation/molten salt method and hydrothermal method.[16,27,28]

YScO:Er3+ and YScO:Er3+,Yb3+ with seven different molar ratios ((mol Sc)/(mol Y + mol Sc) = 0, 0.125, 0.25, 0.375, 0.5, 0.75, and 1) and four different Yb3+ concentrations (5, 10, 15, and 20 mol %) were synthesized using a two-step co-precipitation/molten salt method. For example, the Y1.75Sc0.25O3:Er3+ (5 mol %),Yb3+ (5 mol %) was prepared by dissolving 1.75 mmol of yttrium(III) nitrate hexahydrate (Y(NO3)3·6H2O, Alfa Aesar, 99.9%), 0.25 mmol of scandium(III) chloride hexahydrate (ScCl3·6H2O, Alfa Aesar, 99.9%), 0.1 mmol of erbium(III) chloride hexahydrate (ErCl3·6H2O, Aldrich Chemistry, 99.9%), and 0.1 mmol of ytterbium(III) chloride hexahydrate (YbCl3·6H2O, Aldrich Chemistry, 99.9%) in 50 mL of deionized (DI) water under vigorous magnetic stirring. Next, 50 mL of ammonium hydroxide (NH4OH, 28–30%, ACS grade) was added to the above solution dropwise and stirred for 2 h. The precipitate was filtered and washed with DI water several times until the pH of the solution was 7. After washing and filtering, the precipitate was dried overnight at 100 °C. The dried precipitate was then mixed and grounded with a eutectic mixture of NaNO3 (high purity grade, VWR, 99.0%) and KNO3 (ACS grade, VWR) to form a homogeneous powder. The mixture was heated to 650 °C for 6 h with a heating rate of 10 °C/min. After cooling to room temperature, the resultant powder was washed several times with DI water and dried overnight at 100 °C. Then, the as-synthesized powder samples were annealed at 750 °C for 16 h to obtain YScO:Er3+ NPs. A similar synthetic route was employed to prepare the YScO:Er3+,Yb3+ NPs with varying Sc3+ and Yb3+ doping concentrations.

The hydrothermal method was also used to synthesize the YScO:Er3+ (5 mol %) NPs with five different Sc molar ratios (0, 0.25, 0.5, 0.75, and 1).[28] Taking Y2O3:Er3+ (5 mol %) as an example, 2 mmol of Y(NO3)3·6H2O and 0.01 mmol of ErCl3·6H2O were dissolved in 4 mL deionized water followed by the addition of 14 mL of 0.2 M NaOH solution drop by drop under stirring. After stirring for 1 h, the solution was transferred into a 23 mL Teflon autoclave and subsequently sealed and then heated at 180 °C overnight (16 h). The autoclave was allowed to cool naturally, and then the precipitate was washed with deionized water and centrifuged for several times. Afterward, the powder was dried at 90 °C overnight, placed into a porcelain crucible and annealed at 500 °C for 3 h.[28]

Powder X-ray Diffraction (XRD)

The crystal structure was studied by performing powder XRD using a PANalytical X-ray diffractometer operating at 45 kV and 40 mA. The 2θ radial scan was performed using a Cu Kα (λ = 1.54 Å) radiation source from 5 to 70° with a step size of 0.03°. Rietveld refinement was performed on the resultant diffraction patterns using the GSAS II software[29] for structural verification and phase quantification. Full structural refinement was achieved by performing the procedure outlined in ref (16).

Electron Microscopy

Scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDX) were taken using a JEOL JSM-6701F scanning electron microscope operating at 10 kV. The YScO:Er3+ (5 mol %) NPs powder was sonicated in an ethanol solution and then dispersed on a carbon tape attached to the SEM stage and dried naturally.

Photoluminescence (PL) Spectroscopy

The downconversion photoluminescence measurements were performed on an Edinburgh FLS1000 PL spectrometer equipped with a PMT detector and a 450 W ozone-free Xenon arc lamp as a light source. The powder samples were placed into a quartz spectrophotometer cell (Starna Cells, Inc). The excitation and emission scans were collected with a bandwidth of 3 nm, a dwell time of 0.5 s, and a step size of 1 nm in the measured range of 350–500 (excitation) and 525–750 nm (emission). Judd–Ofelt calculations were performed based on the excitation measurements in the range of 350–550 nm. For upconversion measurements, an MDL-III-980 laser centered at 980 nm with a power of 2500 mW was employed as the excitation source and emission spectra were recorded from 450–750 nm. The system is not equipped with an integrating sphere, preventing any UC quantum efficiency measurements. The lifetime measurements were performed with a microsecond flash lamp (frequency: 25 Hz, 1–2 μs pulse) over the range of 10 ms with a 2 ms delay time, resulting in a 5 μs detector response (2000 channels). The low-temperature measurements were performed using a LINKAM THMS600 temperature-controlled stage, which was added to the spectrometer setup. Liquid nitrogen was used as the cooling agent.

Results and Discussion

Y2O3 and Sc2O3 are commonly synthesized using the hydrothermal route for phase and NPs shape control.[28,30] Thus, hydrothermal synthesis was first applied in this work to prepare YScO:Er3+ (5 mol %) NPs. As shown in Figure S1 (Supporting Information), Y2O3:Er3+ (5 mol %) and Sc2O3:Er3+ (5 mol %) exhibit highly crystallized phases matching the Y2O3 (ICDD 05-4378, space group Ia3) and Sc2O3 (ICDD 05-4385, space group Ia3) standard phases. No noticeable shift is observed in the Y2O3 diffraction peaks with Sc3+ doping. Instead, a weak diffraction peak at 31.5° appears for the Y1.5Sc0.5O3 host. The peak intensity increases and shifts to match pure Sc2O3 with increasing Sc3+ concentrations, indicating generation of Sc2O3 segregation instead of a homogeneously mixed YScO lattice. As such, the co-precipitation/molten salt synthesis method was applied to overcome the segregation, as demonstrated for other complex oxides.[25,27,31] The XRD patterns for the YScO:Er3+ (5 mol %) NPs (Figure a) are readily indexed to cubic Y2O3 and Sc2O3 standards. As expected for mixed metal oxides, the diffraction peaks systematically shift to higher angles with increasing Sc concentration. Furthermore, the lattice parameter decreases linearly from 10.6 (Y2O3) to 9.8 Å (Sc2O3), suggesting a homogenous doping of Sc at all concentrations (Figure b).

Figure 1

(a) XRD patterns of as-synthesized YScO:Er3+ (5 mol %) NPs with Y2O3 and Sc2O3 standards using co-precipitation/molten salt synthesis methods. (b) Lattice parameters change as a function of Sc concentration. (c) TEM image of as-synthesized Y2O3:Er3+ (5 mol %).

At the same time the particle size decreases proportional to Sc3+ concentrations (30.25 nm for Y2O3 to 17.24 nm for Sc2O3) based on the Scherrer equation applied to the (220), (400), (440), and (622) peaks. The size matches well with TEM image (Figure c). The diffraction patterns were also refined to confirm the crystal structure and particle sizes. All four YScO:Er NPs listed in the Table S1 (Supporting Information) were refined using a cubic bixbyite structure with the Ia3 space group, in which Y2O3:Er3+ (5 mol %), Y1.75Sc0.25O3:Er3+ (5 mol %) and YScO3:Er3+ (5 mol %) are refined with the cubic Y2O3 phase,[32] while Sc2O3:Er3+ (5 mol %) is refined with the cubic Sc2O3 phase.[33] The fits were deemed to be good based on the statistical values (Rp and Rwp < 10% and χ2 < 10) coupled with visual confirmation (Figure S2). The fitted lattice parameters show good agreement with the calculated values in Figure b. Additionally, SEM/EDX elemental mapping were applied to visualize the contribution of each cation in Y2O3, YScO3, and Sc2O3 NPs (5 mol % Er3+, Figures S3–S5). As shown in the figures, no Sc3+ or Y3+ clustering is present, confirming the NP homogeneity, allowing for further characterization of the optical properties.

DC and UC spectra were collected to determine the luminescence performance of the as-synthesized YScO:Er3+ (5 mol %) NPs. Figure a shows the DC emission and excitation spectra as a function of Sc3+ concentration (measured at room temperature). All YScO:Er3+ (5 mol %) NPs exhibit typical Er3+ DC emission peaks at 553 nm (2H11/2 → 4I15/2), 564 (4S3/2 → 4I15/2), and 661 nm (4F9/2 → 4I15/2) with the 379 nm excitation.[5] The DC intensity significantly increases (×5) with Sc3+ concentrations up to ∼12 mol % (Figure b). Similar enhancement is detected for the UC luminescence (Figure c). The maximum UC intensity is observed with 12–25 mol % Sc3+ doping, after which the PL nearly completely quenches (Figure d). In addition, Figure S6 shows the DC and UC spectra for lower Sc concentration (<12.5 mol %) doped YScO:Er NPs showing increasing intensities that peak at 12.5 mol %. The CIE color coordinates (Figure S7) show no systematic trend with Sc3+ concentration in YScO:Er3+ (5 mol %) NPs. Overall, the DC and UC spectra improve with lower Sc3+ incorporation, enhancing both radiative transitions and is believed to be due to the Sc ion modifying the crystal field around the luminescence center.

Figure 2

Luminescence performance of YScO:Er3+ (5 mol %) NPs with varying Sc concentrations. (a) PL emission spectra excited at 379 nm and excitation spectra with the emission wavelength of 661 nm. (b) Change of PL emission intensities of the peaks at 564 and 661 nm with increasing Sc concentrations. (c) UC spectra excited with a 980 nm laser. (d) Change of UC intensities of the peaks at 564 and 661 nm with increasing Sc concentrations.

Next, the Judd–Ofelt analysis was applied to quantify the role of Sc3+ doping on the luminescence efficiency using the excitation spectra between 350 and 550 nm and monitoring the 564 nm emission peak.[26] As shown in Figure , the characteristic excitation bands of Er3+ ions originating from the 4I15/2 ground state to higher excited states were readily identified.[34] Six Er3+ excitation bands were selected for the evaluation of the Judd–Ofelt parameters, Ω (t = 2, 4, 6), for the corresponding Er3+ transitions.[26,35,36] It has been reported that Ω2 is closely related to the hypersensitive transitions, which means that larger Ω2 values indicate lower local symmetry.[37] The hypersensitivity of the certain lines in the spectra of RE ions originates from the inhomogeneity of the local crystal environment, with the most striking effect expected for highly polarized, asymmetric environments.[38] The calculated Judd–Ofelt parameters are shown in Table . The values of the Judd–Ofelt parameters for the Y2O3:Er3+ (5 mol %) NPs (Ω2 = 6.71 × 10–20 cm2, Ω4 = 2.76 × 10–20 cm2, Ω2 = 1.54 × 10–20 cm2) are comparable to the literature validating the measurement method.[39] The Ω2 of the YScO:Er3+ (5 mol %) particles peaks at 7.31 × 10–20 cm2 for 37.5 mol % of Sc3+ (Y1.25Sc0.75O3:Er3+ (5 mol %)), agreeing with the DC/UC concentration-dependent measurements. In addition to Ω2, Eu3+ was applied as a symmetry probe to replace the Er3+ and further determine the local symmetry change around the luminescent ion sites (Figure S8). The magnetic dipole (MD, λem = 590 nm) and electric dipole (ED, λem = 611 nm) transitions for Eu3+ are commonly used to quantify the local environment change through the asymmetry ratio (R = IED/IMD).[25] The calculated asymmetry ratios are shown in Table S2 and match well with the Judd–Ofelt analysis, indicating that the Sc incorporation changes the local symmetry in the Er3+ sites, which proves the hypothesis made from the PL results.[37]

Figure 3

PL excitation spectra with the emission wavelength of 564 nm of YScO:Er3+ (5 mol %) NPs with varying Sc concentrations measured at −190 C under a liquid nitrogen environment.

Table 1

Judd–Ofelt Parameters Based on PL Excitation of YScO:Er3+ (5 mol %) NPs at 563 nm Emission Measured under a Liquid Nitrogen Environment

[Sc] (mol %)	Ω2 (× 10–20 cm2)	Ω4 (× 10–20 cm2)	Ω6 (× 10–20 cm2)
0	6.71	2.76	1.54
12.5	7.09	2.45	1.60
25	7.09	2.18	1.71
37.5	7.31	1.69	1.78
50	5.72	4.50	0.91
75	6.10	4.12	1.63
100	5.42	5.99	1.17

Next, low-temperature (77 K) lifetimes (Figure S9) were measured for the YScO:Er3+ NPs and fitted using single exponential decay function to quantify the luminescence efficiency (Table ).[40] The measured lifetimes of emissions at 553 nm (2H11/2 → 4I15/2) and 564 nm (4S3/2 → 4I15/2) peaks with the addition of 12.5 mol % of Sc3+ also show a maximum between 12.5 and 25 mol % Sc incorporation, matching the concentration window of seen for the UC and DC intensity measurements. As with the other luminescence measurements, the 2H11/2 → 4I15/2 and 4S3/2 → 4I15/2 transitions also show a maximum between 12.5 and 25 mol % Sc3+ incorporation. For the Y2O3:Er3+ (5 mol %) NPs, the photons needed (n) to excite the 2H11/2 and 4S3/2 energy levels is expected to be 2.[41] In the case of this work, the pump power-dependent intensity of the as-synthesized YScO:Er3+ (5 mol %) NPs (Figure S10) are not linear. This can be explained by the mechanism of excited state absorption (ESA) and energy transfer UC (ETU). In the case of ESA as the predominant UC mechanism, the number of photons (n) is almost independent of pump power.[42] This non-linear P–I curve suggests that Sc3+ results in a breakdown of the symmetry, negatively impacting the energy transfer between cations. Overall, with less than 30 mol % of Sc3+ incorporation in the Y2O3:Er3+ (5 mol %) NPs, the luminescence properties improved due to local symmetry modifications.

Table 2

Luminescence Efficiency Based on Judd–Ofelt Calculation and Decay Lifetime of the Transitions 2H11/2-4I15/2 (553 nm), 4S3/2-4I15/2 (564 nm), and 4F9/2-4I15/2 (661 nm) with the 379 nm Excitation of YScO:Er3+ (5 mol % ) NPs Measured under a Liquid Nitrogen Environment

[Sc] (mol %)	τcalc (μs)	τexp (μs)	efficiency	τcalc (μs)	τexp (μs)	efficiency	τcalc (μs)	τexp (μs)	efficiency
0	57	32	56%	265	30	11%	226	42	19%
12.5	56	42	75%	256	39	15%	242	36	15%
25	57	39	69%	240	37	15%	254	35	14%
37.5	57	34	59%	232	31	13%	289	29	10%
50	57	25	44%	438	24	5%	173	18	10%
75	56	20	36%	250	20	8%	167	46	28%
100	53	15	28%	338	16	5%	131	10	8%

As a comparison, the Yb3+ ion was co-doped into the YScO:Er3+ (5 mol %) NPs as a sensitizer using the previously described co-precipitation/molten salt process (Figure S11). Figure shows the UC spectra of the Y1.75Sc0.25O3:Er3+ (5 mol %),Yb3+ (0–20 mol %) NPs. Comparison of the UC intensities (insert of Figure a) shows that the 5 mol % Yb3+ resulted in the highest sensitization for the 661 nm (4F9/2 → 4I15/2) emission. Further increasing the Yb doping to 20 mol %[41] resulted in significant quenching. Additionally, the red to green ratio (I661nm/I563nm) increases with increasing Yb3+ concentration (Figure b) as a result of enhanced Yb3+-Er3+ energy transfer. This trend is as expected as the distance between RE dopants is inversely proportional to the concentration but also results in enhanced surface quenching (lower intensities). On top of this, it is believed that the larger Yb3+ ions (0.87 Å) can expand the unit cell, negating the impact of the smaller Sc3+ ions. The power-dependent UC intensity was measured for different Yb3+ concentrations (Figure c). As opposed to the single Er doped system, the co-doped system exhibited a linear power dependency indicative of efficient Yb3+-Er3+ energy transfer. Interestingly, the number of required photons (n) increases proportionally with Yb3+ doping, again suggesting increased energy transfer and quenching due to parasitic surface groups. The UC mechanism is illustrated with the energy level diagram shown in Figure . As Er3+ is the only dopant in the YScO host, the UC process occurs through ground-state absorption (GSA) followed by excited-state absorption (ESA). The excited electrons relax to a stable state via non-radiative relaxation followed by radiative relaxation (Figure a). Since GSA and ESA are both low efficiency excitation processes, Yb3+ was introduced into the system due to its large absorption cross section (∼10–20 cm2).[43,44] In this case, the GSA and ESA processes are diminished, while tge energy transfer (ET) process becomes the dominant excitation mechanism for Er3+, improving the overall UC efficiency (Figure b).[45]

Figure 4

(a) UC spectra, (b) red/green ratio, and (c) pump power dependence of the 661 nm emission intensity of the of Y1.75Sc0.25O3:Er3+ (5 mol %),Yb3+ (x mol%, x = 0, 5, 10, 15, 20) NPs.

Figure 5

Energy level diagram and energy-transfer mechanism of (a) Er3+-doped and (b) Er3+/Yb3+-co-doped YScO NPs under 980 nm excitation.

Furthermore, to calculate the efficiency of YScO:Er3+,Yb3+ NPs, the Judd–Ofelt analysis was also applied (Figure S12a and Table ). The Ω2 value increases up to 10 mol % Yb3+, indicating further amplification of the hypersensitive transitions (and decreased symmetry) despite the luminescence quenching. To further quantify the luminescence efficiency, lifetimes for the transitions of interest all decreased with increasing Yb3+ concentration (Table and Figure S12b–d). Both the experimental and theoretical lifetimes (and efficiency) slightly decreased for the 5 mol % Yb3+ sample compared to the samples without Yb3+. However, higher dopant concentrations significantly impacted the luminescence efficiency. Based on the slate of PL characterization measurements, the 5 mol % Yb3+ was deemed the optimal doping concentration to sensitize the luminescence.

Table 3

Judd–Ofelt Parameters Based on PL Excitation of Y1.75Sc0.25O3:Er3+ (5 mol %),Yb3+ (x mol%, x = 0, 5, 10, 15, 20) NPs at 563 nm Emission Measured under a Liquid Nitrogen Environment

[Yb] (mol %)	Ω2 (× 10–20 cm2)	Ω4 (× 10–20 cm2)	Ω6 (× 10–20 cm2)
0	7.43	1.87	1.59
5	7.96	2.19	1.43
10	8.12	1.76	1.54
15	7.82	1.65	1.83

Table 4

Luminescence Efficiency Based on Judd–Ofelt Calculation and Decay Lifetime of the Transitions 2H11/2-4I15/2 (553 nm), 4S3/2-4I15/2 (564 nm), and 4F9/2-4I15/2 (661 nm) with Excitation at 379 nm of the Y1.75Sc0.25O3:Er3+ (5 mol %), Yb3+ (x mol%, x = 0, 5, 10, 15, 20) NPs Measured under a Liquid Nitrogen Environment

	553 nm			563 nm			661 nm
[Yb] (mol%)	τcalc (μs)	τexp (μs)	efficiency	τcalc (μs)	τexp (μs)	efficiency	τcalc (μs)	τexp (μs)	efficiency
0	56	42	75%	259	39	15%	286	36	13%
5	52	37	71%	287	38	13%	270	30	11%
10	52	19	36%	267	19	7%	300	19	6%
15	54	15	27%	225	13	6%	289	14	5%

Based on the luminescence measurements above, low Sc3+ concentrations in the Y2O3:Er3+ NPs improve the Er3+ UC and DC intensities by breaking the local symmetries around these sites. By adjusting the size of cation sites with a certain amount of smaller Sc3+ ions, the local symmetry around the RE ions is reduced. Interestingly, the phosphors with the highest intensity and luminescence efficiencies had an Y:Sc ratio of ∼3:1, similar to the C2 and C3 site amount ratio, suggesting that Sc may preferentially occupy the C3 sites.

Conclusions

YScO:Er3+ and YScO:Er3+,Yb3+ NPs were synthesized using co-precipitation and molten-salt synthesis methods. As a comparison to the hydrothermal method, XRD patterns showed a systematic shift of the diffraction peaks and linear change in the lattice parameters with increasing Sc concentrations, indicating that Sc can be incorporated into Y2O3 lattices and form homogeneous nanocrystals. Sc incorporation was able to improve the overall luminescence performances of Er in the YScO host. Specifically, UC and DC spectra showed that less than 30% of Sc3+ incorporation significantly improved the luminescence intensities, in which Y1.75Sc0.25O3:Er3+ (5 mol %) NPs exhibited the strongest UC and DC emissions. Furthermore, the comparison between theoretical (Judd–Ofelt analysis) and experimental (lifetimes measurements) methods were used to determine the luminescence efficiency. The calculated Judd–Ofelt parameter Ω2 showed that Sc3+ substitution of Y3+ in Y2O3 improved the local asymmetry level around the Er3+ sites, while Ω2 increased with less than 50 mol % of Sc3+ and reduced with more Sc3+ incorporation. A similar trend was obtained in the fitted lifetime and calculated efficiency, showing that Y1.75Sc0.25O3:Er3+ (5 mol %) NPs had the longest lifetime and the highest efficiency and less than 30 mol % of Sc3+ is the concentration window, which improved the luminescence performance. Next, Yb3+ was introduced into YScO:Er3+ as a sensitizer to further optimize the luminescence performance. It was found that Sc3+ substitution significantly reduced the Yb3+ concentration to a 1:1 Yb3+:Er3+ ratio. Meanwhile, Sc3+ substitution showed the same effect on the YScO:Er3+,Yb3+ NPs as to the YScO:Er3+ NPs where the Y1.75Sc0.25O3:Er3+ (5 mol %),Yb3+ (5 mol %) NPs exhibited the highest UC and DC luminescence intensities and overall efficiency. Overall, this work offers an idea to spatially control the doping of luminescent RE ions into more asymmetric cation sites to obtain a better luminescence performance, which is required for many applications such as solid-state phosphors, optical thermometers, and luminescence probes.