Red-emitting BaAl₂O₄:Eu³⁺ synthesized via Pechini and sol–gel routes: a comparison of luminescence and structure

Abstract

UV-to-red downshifting phosphors such as BaAl₂O₄:Eu³⁺ find broad range of application in sensors, displays, and in solid-state lighting, yet new synthetic routes to improve their luminescence are envisaged. In this regard, herein, it is introduced two new methods to synthesize this environmentally friendly BaAl₂O₄:Eu³⁺, by an adapted sol–gel route and a modified Pechini synthesis. Additionally, a systematic study was carried out about the Eu³⁺ doping concentration and charge compensation effects on the structural, morphological and spectroscopic features. Both routes enabled high-crystalline and nanostructured phosphors displaying optic bandgap near to 4.4 eV, although the sol–gel route also led to low amounts of BaCO₃ spurious phase. Upon UV (250 nm) excitation, all Eu³⁺-doped samples emit red light displaying high emission color purity, characteristic of the ⁵D₀ → ⁷F_0-4 electronic transitions of Eu³⁺. The Pechini method led to the highest intrinsic emission quantum yield (85% for the 3%-doped sample). Eu³⁺ replaces Ba²⁺ within the BaAl₂O₄ lattice, but in the sol–gel-derived samples, the dopant may also replace Ba²⁺ into the BaCO₃ spurious phase, confirming that the Pechini route is the best one to optimize the luminescence and structure of the phosphor.

Graphical abstract

Introduction

UV-to-red downshifting emitting phosphors find a broad range of applications [1] in televisions [2, 3], solid-state lighting [4, 5], multifunctional light-emitting diodes [6, 7], field emission displays [2, 3], photonics, plasma display panels (PDPs) [2, 3], fiber amplifiers [8], optoelectronics for image storage [9], and detectors of energy radiation [9]. Although the main commercially available red-emitting phosphors, i.e., Zn_1-xCd_xS:Ag⁺ or Y₂O₂S:Eu³⁺, feature high overall emission quantum yield, some drawbacks must still be overcome, especially that related to their composition based on heavy metals and sulfur atom that is released over time of usage [3, 10]. In this regard, new strategies toward the synthesis of environmentally friendly red-emitting phosphors displaying intense luminescence are of large societal relevance.

Among several red-emitting systems, BaAl₂O₄:Eu³⁺ stands out due to its environmentally friendly properties, chemical, thermal, and photostability, relatively low cost of production, simplicity of the synthesis methods, and similarity between the ionic radius of Ba²⁺ and Eu³⁺, altogether making BaAl₂O₄ a desirable host. [11] Moreover, Eu³⁺ is a well-known luminescent activator characterized by narrow emission bands within the red spectral region that arise from intraconfigurational 4f → 4f electronic transitions [12]. Due to these features, several studies have shown the potential application of BaAl₂O₄:Eu³⁺ in solid-state lighting, scintillation, or dosimetry [13, 14]. Yet, some issues need to be addressed concerning the development of new synthetic strategies toward the enhancement of the Eu³⁺ luminescence and the charge imbalance between Eu³⁺ and Ba²⁺, which leads to structural defects that often quench the radiative process arising from the ⁵D₀ emitting state of Eu³⁺. [15]

Some routes for the obtention of BaAl₂O₄:Eu³⁺ are reported so far, such as methods based on combustion [13, 16], conventional Pechini [11, 15], hydrothermal [17], floating zone [13], and protein [18,19,20,21] or polyvinyl alcohol [22]-assisted sol–gel. Moreover, methods applied for the synthesis of other Eu³⁺-doped aluminates are also reported, such as the combustion method used in the ZnAl₂O₄:Eu³⁺:Tb³⁺ synthesis [23] or the urea-assisted combustion approach for obtaining CaAl₂O₄:Eu³⁺,Er³⁺ [24]. Although some of these routes enable a quite good particle shape and size control, there is not a consensus on which synthesis is the best regarding the enhancement of the Eu³⁺ luminescence. In this regard, we introduce two new routes for the phosphor obtention based on a sol–gel route adapted from the Ba₂SiO₄:Eu³⁺ synthesis [25] and a modified Pechini method using sorbitol instead of the classic ethylene glycol as polymerizing agent, which enhances the Eu³⁺ luminescence according to our previous study for the Y₂O₃:Eu³⁺,Er³⁺,Yb³⁺ phosphor [26]. The justifications behind these selections lie in the relatively low cost, simplicity, achievement of high level of purity, well-crystallized phosphor, and decrease in the annealing temperature compared to solid-state synthesis [27, 28].

Therefore, in this study, BaAl₂O₄:Eu³⁺ red-emitting phosphors were synthesized by two adapted routes, the sol–gel and the Pechini method, by applying a charge compensation mechanism in the Eu³⁺ doping within the BaAl₂O₄ lattice. Furthermore, structural, morphological, and spectroscopic studies were performed to evaluate Eu³⁺ site occupancy, influence of concentration on structure and luminescence as well as the mechanism behind energy transfer among Eu³⁺ ions, which is responsible for luminescence quenching by concentration.

Materials and methods

Acetic acid (CH₃COOH, Cinética, 99.7%), citric acid (C₆H₈O₇, Fluka analytical, 99.5%), nitric acid (HNO₃, FMaia, 65%), barium acetate (Ba(CH₃COO)₂, Vetek, 99%) aluminum nitrate (Al(NO₃)₃.9H₂O, Across, 98%), D-sorbitol (C₆H₁₄O₆, Aldrich, 99%), aluminum isopropoxide (C₉H₂₁O₃Al, Across, 99%), and europium oxide (Eu₂O₃, Aldrich, 99.99%) were used as starting reactants without any further purification.

Modified Pechini synthesis. Stoichiometric amount of Ba(CH₃COO)₂ aiming to prepare 0.5000 g of the final product was dissolved in acetic acid (6 mL) under agitation and heating (50 °C), followed by the addition of H₂O (25 mL). After 10 min, stoichiometric amounts of Al(NO₃)₃, Eu(NO₃)₃ (prepared from the dissolution of Eu₂O₃ in concentrated nitric acid, in the case of the doped samples, and titrated by using EDTA as complexing agent and xylenol orange as indicator), and citric acid (3 mols of citric acid:1 mol of metal) were added, enabling the metal-citrate complex formation. The temperature of the system was kept at ~ 85 °C/1 h until halving the initial volume, and then, sorbitol was added (1 mol of sorbitol:2 mols of metal). The system was kept under stirring and heating until the formation of a clear foamy polymer resin, which was pre-calcined (300 °C) for 2 h in static air. A spongy-like agglomerate, also called “puff,” was then obtained [26]. The “puff” precursor was de-agglomerated and sieved at 325 mesh sieves, and annealed at 1000 °C (heating ramp of 5 °C/min) for 2 h in a circulating air atmosphere.

Sol–gel synthesis. Stoichiometric amount of Ba(CH₃COO)₂ aiming to prepare 0.5000 g of the final product was dissolved in acetic acid (1.5 mL) under agitation and heating (50 °C), followed by the addition of H₂O (10 mL). After 10 min, stoichiometric quantities of Eu(CH₃COO)₃ solution (previously prepared from the dissolution of Eu₂O₃ in CH₃COOH) was added. In the sequence, C₉H₂₁O₃Al in stoichiometric proportions was added under vigorous stirring and heating (80 °C) until complete dissolution and gel formation. This gel was dried in an oven for 1 h to yield the xerogel precursor that was thermally treated at 300 °C in a static air atmosphere (heating ramp of 5 °C/min for 2 h) [25]. Finally, the precursor was de-agglomerated in a mortar and annealed at 1,000 °C in a circulating air atmosphere (heating ramp of 5 °C/min) for 2 h.

The Eu³⁺ doping concentrations were isoelectronically varied from 1 to 7% (Table 1), i.e., keeping the same overall charge in the BaAl₂O₄ matrix, according to the Ba²⁺, Eu³⁺ and $Al_{2} O_{4}^{2 - }$ charges ($3n_{{Eu^{3 + } }} + 2n_{{Ba^{2 + } }} = + 2$, where $n_{{Ba^{2 + } }}$ and $n_{{Eu^{3 + } }}$ are the mol number of Ba²⁺ and Eu³⁺, respectively, in the Ba_nBaEu_nEuAl₂O₄ formula). Samples will be hereafter labeled as PCX or SGX, where SG = sol–gel-derived samples, PC = Pechini-derived samples, and X = Eu³⁺-doping percentage.

Table 1 Doping proportions. The Eu³⁺ concentration was isoelectronically varied from 1 to 7%,

Characterization. All samples were characterized, from the structural point of view by X-ray diffraction (XRD) using a SHIMADZU diffractometer model XDR-6000, Cu(K_α) radiation (λ = 1.5406 Å), interval of 10^o—80º (2θ), increment of 0.02º and integration time of 0.6 s. The crystallite size of samples was evaluated via Scherrer method, according to Eq. 1 [29], where βp is the width at half maximum in radians of a given (h k l) reflection peak due to the grain size (ε) in the corresponding (h k l) direction, and c is a constant that depends on the reflection symmetry which usually is given as 1. To calculate the crystallite size, the three most intense peaks related to (2 0 0), (2 0 2), and (2 2 0) plans at 19.7°, 28.2° and 34.2° were considered, respectively:

$$\beta_{p} = \frac{c\lambda }{{\varepsilon {\text{cos}}\theta }}$$

(1)

Thermogravimetry of precursors doped with 1% was measured by using an SDT-Q600TA instrument with a heating rate of 10 °C/min from 25 °C to 1,400 °C in air flow (100 mL/min). SEM measurements were carried out using a scanning electron microscope CARLS ZEISS, model EVO LS15 with a secondary electron detector (SE), 25.00 KX of magnification, EHT of 30.00 kV. For that, an ethanolic suspension containing particles was deposited on a glass cover plate, which was fixed with double carbon facing tape on a stub and then metalized with gold using a Quorum sputtering, model Q 150R ES.

UV–Vis diffuse reflectance was measured in a PERKIN ELMER LAMBDA 1050 spectrophotometer. From the diffuse reflectance spectrum, the optical bandgap was estimated through a graphic of (αhν)ⁿ versus the energy of the incident photon (eV), Figure S1 (Supplementary Material), where n is equal to 2 for direct electronic transitions or 0.5 for indirect transitions between the valence and conduction bands [30]. α is the ratio of the scattering (K) and absorption (S) coefficients determined from the Kubelka–Munk method [31], Eq. 2, where R is the reflectance observed for different incident energies.

$${\upalpha } = { }\frac{{\text{K}}}{{\text{S}}} = \frac{{\left( {1 - {\text{R}}} \right)^{2} }}{{4{\text{R}}}}$$

(2)

Photoluminescence measurements at 300 K or at 77 K were carried out in a Horiba Jobin Yvon spectrometer Fluorolog-3 with a Xe lamp (450 W) source with double excitation monochromator. The slits were placed at 1.1 and 0.7 mm for excitation and emission, respectively. The emission decay curves were carried out in a phosphorimeter equipped with Xe (5 J/pulse) lamp. By applying the Judd–Ofelt theory that describes the behavior of f-f electronic transitions in Ln³⁺ by a forced electric dipole mechanism, it is possible to infer on the radiative decay probabilities (A_rad) of the ⁵D₀ emitting state and the Ω₂ and the Ω₄ intensity parameters. A_rad was calculated by using Eq. 3, where A₀₁ = 14.65 n³ in s⁻¹ [32], and n is the refractive index, equal to 1.657 for BaAl₂O₄ [33], I is the integrated area under the ⁵D₀ → ⁷F_J transitions in the emission spectra, and hυ_0-J is the energy barycenter of the electronic transition. Thus, A_rad is the sum of radiative rates of all electronic transitions coming from the ⁵D₀ state (A_rad = $\sum A_{0 - J}$).

$$A_{0 - \lambda } = A_{0 - J} = A_{01} \frac{{I_{0 - J} }}{{I_{0 - 1} }}\frac{{hv_{0 - 1} }}{{hv_{0 - J} }}$$

(3)

The intensity parameters were calculated by using Eq. 4, where $\left| {{}_{{}}^{7} F_{J} U^{\left( \lambda \right)} {}_{{}}^{5} D_{0} } \right|^{2}$ represents the square reduced matrix elements whose values are equal to 0.0032 to Ω₂ and 0.0023 to Ω₄; h is the Planck's constant, e is the electron charge, c is the speed of light in vacuum, ω is the angular frequency of the incident radiation field, ε₀ is vacuum permittivity constants, and X is the Lorentz local field correction equal to n(n² + 2)²/9 [34].

$$\Omega _{\lambda } = ~\frac{{3hc^{3} A_{{0\lambda }} }}{{8\pi e^{2} \omega ^{3} \chi \left| {\left\langle {^{7} F_{J} \left\| {U^{{\left( \lambda \right)}} } \right\|^{5} D_{0} } \right\rangle } \right|^{2} }}~$$

(4)

Results and discussion

Structural characterization

Precursors of both methods were characterized by thermogravimetry aiming the determination of the BaAl₂O₄ phase formation temperature, Fig. 1. For the sol–gel-derived precursor, the thermal event starting at 200 °C and peaking at 280 °C is correlated with the thermal decomposition of the organic matter coming from the precursors. On the other hand, for the Pechini precursor, the first thermal event with maximum at 100 °C is due to water evaporation, while the second one at 200 °C is related to the thermal degradation of nitrates and organic matter. Due to its high basicity, Ba²⁺ tends to form BaCO₃ in such environmental conditions, whose formation ranges within 400–500 °C [35]. For both precursors, the thermal event at 900 °C is associated with the BaAl₂O₄ phase formation, which was confirmed by XRD data.

Figure 1

Thermogravimetric curves (TG) and first-order derivate (DTG) of a sol–gel and b Pechini precursors, both doped with 1% of Eu³⁺

The X-ray diffraction patterns of samples synthesized at 1,000 °C, Fig. 2, are indexed as the hexagonal BaAl₂O₄ structure (JCPDS-17–0306), space group P6₃22. All diffraction peaks are sharp and well defined, confirming that the phase is highly crystalline. However, for the sol–gel-derived samples with low Eu³⁺ doping percentage, a low-intensity peak at 24° is observed, characteristic of the diffraction plan of the BaCO₃ (JCPDS-45–1471) spurious phase (see the magnification of this region in supplementary material, Figure S2). The BaCO₃ phase formation is common for compounds with high basicity cations such as Ba²⁺, being intermediated by CO₂ coming from organic matter decomposition and/or air atmosphere, as reported for the synthesis of BaTiO₃ [36, 37] or Ba₂SiO₄ [25]. The sol–gel route itself normally leads to particles displaying porous surface that can easily react with CO₂ from atmosphere, favoring the formation of carbonate groups. Moreover, the sol–gel-derived xerogel precursor based on Si–O bonds is often reported [38,39,40] as a porous structure that can confine organic groups, making it difficult for CO₂ to be released due to thermolysis, which favors the BaCO₃ formation. Interestingly, it is observed a smooth peak at 24° in the XRD data reported by Chatterjee and co-workers for sol–gel-derived BaAl₂O₄ [15], which may be assigned to the BaCO₃ phase.

Figure 2

X-ray diffraction pattern of the Eu³⁺-doped BaAl₂O₄ samples obtained from a sol–gel and b Pechini routes. c BaAl₂O₄ crystalline structure based on the CIF file available on Inorganic Crystal Structure Database (ICSD)

In BaAl₂O₄, Ba²⁺ occupies two different sites with the same coordination number (CN = 9), Fig. 2c, but in the Ba1 site (C₃ point group), the Ba–O bond distance lies within the 2.86 – 2.87 Å, while for the Ba2 site (C₁ point group), the Ba–O bond distance is 2.69 Å. Moreover, the Ba1 site occurs at about three times more than the Ba2 one. On the other hand, Al³⁺ is inserted into four different tetrahedral sites featuring Al-O bond distances close to 1.77, 1.74, 1.72, and 1.83 Å. [9] Since Eu³⁺ (ionic radius of 126 pm, CN = 9) and Ba²⁺ (ionic radius of 161 pm) have a better ionic radii match, Eu³⁺ replaces Ba²⁺ sites in the BaAl₂O₄ structure according to the Hume-Rothery rules, which states that the ionic radius difference between solvent and solute ions should be close to 15% [9]. It is worth pointing out that although the ionic radii difference between Eu³⁺ and Ba²⁺ is larger than 15%, doping takes place since no Eu₂O₃ diffraction peaks are observed in the XRD diffractogram (considering the detection limit of 1wt.%), Fig. 2, confirming the formation of a solid solution. Interestingly, Ba²⁺ replacement by lanthanide(III) in inorganic matrices has already been reported, for instance, for the Ba₂SiO₄:Eu³⁺ [25] and Ba₂SiO₄:Tb³⁺ [4] systems.

From the analysis of Fig. 2, there is no shift of the diffraction peaks related to the BaAl₂O₄ phase in the XRD data for the Pechini-derived samples (Figure S2), suggesting that neither the doping nor the synthetic route led to significant changes in the BaAl₂O₄ crystalline structure. Yet, for the sol–gel-prepared ones, slight shifts of the diffraction peaks (Figure S2) may be correlated with the spurious phase that influences the BaAl₂O₄ lattice organization.

The crystallite size values determined by the Scherrer equation shown in Table 2 are comparable with those ones of other oxide compounds prepared by humid routes [35, 41, 42] and they also reveal a non-isotropic crystallite grown tendency. There is a tendency of increasing the average crystallite size due to the doping for the SG samples, while the values for the Pechini-derived ones have no statistical change. In this case, changes in the crystallite size may be correlated with two factors, the BaCO₃ spurious phase (present only in the SG1 and SG3 samples) and the doping amount. The amount of BaCO₃ phase may control the boundaries during the synthesis, determining the mass transfer through the grain boundaries [43]. On the other hand, since Eu³⁺ and Ba²⁺ have quite different ionic radii, the increase in the doping content may lead to smooth local changes around Eu³⁺ local sites, which influence the structure formation and as a consequence the crystallite size.

Table 2 Crystallite size (± 1 nm) for the three most intense diffraction peaks and average crystallite size for samples

Representative SEM images of PC1 and PC5 samples, Fig. 3a and b, exhibit particle agglomerates, which are common to quite high-temperature annealing processes, resulting in sintered surfaces [44]. On the other hand, the sol–gel-derived particles, Fig. 3c and d, feature different shapes and sizes, with spheroidal-like particles on the surface. This behavior is associated with the acid-catalyzed sol–gel route, which typically leads to particle agglomerates from the coalescence of spheres [45]. Thus, the sol–gel synthesis enables better shape control, although both methods form a nanostructured bulk.

Figure 3

SEM images of a PC1, b PC5, c SG1, and d SG3 samples

Optical band gap evaluation

The optical absorption of all samples was evaluated through diffuse reflectance spectroscopy, Fig. 4. For all doped samples, three bands are noticed in the diffuse reflectance spectra; the first one at 220 nm (also noticed in the excitation spectra, Fig. 5) is assigned to the Eu³⁺ → O²⁻ charge transfer band (CTB), and its relative intensity enhances as the Eu³⁺ content increases; the second at 262 nm is the electronic transition from the valence band (VB) to the conduction band (CB) of the matrix (the optical bandgap), and the third at 361 nm, also observed in the non-doped samples, may be associated with absorptions to electronic states arising from structural defects.

Figure 4

Diffuse reflectance spectra of a sol–gel and b Pechini-derived samples. c Energy diagram representing the absorption transitions. Assignments: (1) Eu³⁺ → O²⁻ charge transfer band (CTB). (2) Valence band (VB) → conduction band (CB) transition. (3) Defect-related matrix absorption

Figure 5

Excitation spectra of the a sol–gel and b Pechini-derived samples monitored at 616.7 nm

Indeed, it is usual to notice absorption in inorganic matrices ranging in the near-UV spectral region due to the presence of traps or intrinsic defects that may arise from the thermal decomposition of organic solvents or organic products of the sol–gel or Pechini syntheses [25, 35]. Moreover, due to the isoelectronic doping, three Ba²⁺ sites are replaced by only two Eu³⁺ ions, leading to one Ba²⁺ vacancy according to Eq. 5. Those vacancies increase the concentration of structural defects, leading to the appearance of defect-related electronic states within the forbidden band, as represented in Fig. 4c. For sol–gel-derived samples, the higher is the doping percentage, the stronger is the defect-related absorption band at 350 nm, ensuring that the structural defect concentration increases. On the other hand, for Pechini-based samples, this behavior is not observed due to the larger amount of organic matter within the precursor coming from sorbitol and citric acid, which also induces structural defects, decreasing the importance of defects arising from the doping process.

$$3Ba^{2 + } + 2Eu^{3 + } \to V_{Ba} + 2Eu_{Ba}^{ \cdot }$$

(5)

The optical bandgap of samples was calculated accordingly to the Kubelka–Munk approximation, Figure S1, and the values are statistically the same (4.3 ± 0.1 eV) independent on the doping concentration or the synthesis, in accordance with other values reported so far for the BaAl₂O₄ host [46].

Luminescence

The excitation spectra of samples monitored at emission of Eu³⁺, Fig. 5, are dominated by a broad and intense excitation band within the short-wavelength region assigned to the Laporte’s allowed Eu³⁺ → O²⁻ CTB, in accordance with reflectance diffuse spectra, confirming that the BaAl₂O₄ host sensitizes the Eu³⁺ luminescence. Moreover, the lower-intensity excitation lines ranging in the long-wavelength region are attributed to the Laporte’s forbidden intraconfigurational Eu³⁺ f-f transitions [47]. In this sense, the phosphors may be excited from the deep UV to the blue spectral window, crossing the near UV spectral region.

The emission spectra of all phosphors, Fig. 6, are characterized by a set of sharp emission bands within the orange-red spectral region assigned to electronic transitions coming from the ⁵D₀ state of Eu³⁺ to ⁷F_0-4 ones, where the band peaking at 617 nm assigned to the ⁵D₀ → ⁷F₂ transition is the most intense one [47]. Interestingly, the Pechini-derived phosphors also feature a broad and weak emission band at 420 nm, which is assigned to the matrix emission accordingly to the emission spectrum of the undoped BaAl₂O₄ (see Figure S3); this band arises from the aforementioned defect-related electronic states according to the diffuse reflectance data and its stronger intensity in the Pechini-derived phosphors is related to the higher concentration of organic matter in the Pechini synthesis, which may induce a higher concentration of structural defects in the matrix.

Figure 6

Emission spectra of a sol–gel and b Pechini-derived phosphors excited at 250 nm. 1,931 CIE chromaticity diagram of c sol–gel and d Pechini-derived phosphors

From the colorimetric point of view, the emission color of phosphors was characterized by the 1,931 Commission Internationale de l'éclairage (CIE) color coordinates, Fig. 6 and Table S1. The emission color of all sol–gel-derived phosphors lies within the red spectral region with color purity near to 100% although, for the Pechini-derived samples, there is a shift toward the pink due to the combination of the weak matrix blue luminescence and the Eu³⁺ characteristic red emission.

Besides the red luminescence, Eu³⁺ may act as a spectroscopic probe since some of its f-f electronic transitions are influenced by the ligand field, enabling the evaluation of the local Eu³⁺ microssimetry by the relative intensity and the number of components of emission bands [48]. In the emission spectra shown in Fig. 6, the most intense band arises from the ⁵D₀ → ⁷F₂ transition, which occurs by a forced electric dipole mechanism, suggesting that Eu³⁺ is inserted in a low-symmetric site, in accordance with Eu³⁺ replacing Ba²⁺ local sites (C₁ or C₃ point group) within the BaAl₂O₄ lattice.

Eu³⁺ also enables a deep evaluation of the number of non-equivalent Eu³⁺ local sites with C_nv, C_n, and C_s point group since the ⁵D₀ state is non-degenerated, leading to only one component for the ⁵D₀ → ⁷F₀ band for each non-equivalent Eu³⁺ local site according to the J rule (2 J + 1) [49]. To get further insight on the Eu³⁺ local microssimmetry, low-temperature (77 K) selective excitation spectra were carried out in the range of the ⁵D₀ → ⁷F₀ band for representative SG1, SG5, PC1, and PC5 phosphors, Fig. 7. PC1 and PC5 feature 1 and 2 ⁵D₀ → ⁷F₀ transition components, respectively, suggesting that Eu³⁺ is inserted in at least two non-equivalent Eu³⁺ local sites, in accordance with Eu³⁺ replacing both Ba²⁺ sites in the BaAl₂O₄ lattice. On the other hand, SG1 and SG5 display 2 and 3 ⁵D₀ → ⁷F₀ transition components, respectively, confirming that Eu³⁺ is inserted into 2 and 3 sites lacking inversion center. In this case, it must be highlighted that Eu³⁺ may replace Ba²⁺ into the BaCO₃ spurious phase found in low amounts in the sol–gel-derived samples.

Figure 7

Selective excitation spectra (77 K, changing the emission wavelength, λ_em) within the ⁵D₀ → ⁷F₀ transition range, monitoring the components of the ⁵D₀ → ⁷F₂ transition (λ_em) for the a PC1, b PC5, c SG1 and d SG5 samples

Based on the Judd–Ofelt theory that describes the f-f emission of Ln³⁺ by the forced electric dipole mechanism, it is possible to determine Ω₂ and Ω₄ intensity parameters, Table 3. A study carried out by Malta and coworkers [50] has shown that the Ω₂ parameter is more influenced by angular distortions of the Eu³⁺-ligand bond, in other words, its local microssymmetry, while Ω₄ suffers stronger influence of the Eu³⁺-ligand bond polarizability.

Table 3 Judd–Ofelt Ω₂ (10^–20 cm⁻²) and Ω₄ (10^–20 cm⁻²) intensity parameters, ⁵D₀ lifetime (τ / ± 0.01 ms) and average lifetime (< τ > / ± 0.01 ms), ⁵D₀ state radiative (A_rad / s⁻¹) and non-radiative (A_nrad / s⁻¹) decay probabilities, and intrinsic emission quantum yield ($\phi_{Ln}^{Ln}$). The Eu³⁺-Eu³⁺ critical distance (R_c, Å) is also shown

The quite low Ω₄ values for samples reveal a predominant ionic nature of the Eu³⁺–O²⁻ bonds, as expected since the Ba²⁺-O²⁻ bonds are mainly ionic. For the Pechini-derived samples, the Ω₄ parameter increases from PC1 to PC3 and then, it decreases, suggesting an increase in the bond covalent degree up to the doping concentration of 3%. On the other hand, for the samples prepared by the sol–gel route, except for the SG3, the values are practically invariant; in these samples, it must be highlighted that the BaCO₃ spurious phase plays an important role in the crystal growth as discussed by XRD, influencing the organization of the BaAl₂O₄ lattice.

The Ω₂ values are close for all Pechini-derived samples, suggesting that the Eu³⁺ concentration does not lead to significant changes on the Eu³⁺ local microssimetry, in accordance with the XRD data. However, samples synthesized by sol–gel display significant changes in the Ω₂ parameter as the Eu³⁺ amount increases, especially the SG3 sample. Again, in this case, the spurious phase may influence the network organization during the synthesis. As observed by the XRD data, Figure S2(a), the relative intensity of the peak at 24° assigned to BaCO₃ is not the same for all samples, suggesting that the weight fraction in the mixture is different. In this context, for the SG3 sample, the BaCO₃ weight fraction should be quite distinct from the others, causing deviations of the typical crystalline structure, especially modifications of Eu – O bond angles whose behavior influences the Ω₂ value.

The ⁵D₀ state lifetime values, Table 3, were obtained through a biexponential adjustment of the emission-decay curves shown in Figure S4. In this case, two ⁵D₀ state lifetime values were found and they were used to calculate an average lifetime by applying Eq. 6, where τ is the ⁵D₀ state lifetimes and A is a pre-exponential value obtained from the biexponential adjustment. The main importance of the lifetime value is the balance between the non-radiative (A_nrad) and radiative (A_rad) decay probabilities from the ⁵D₀ emitting state (1/τ = A_rad + A_nrad). Thus, apart from the ⁵D₀ state lifetime value and the A_rad determined from Eq. 3, the intrinsic emission quantum yield of Eu³⁺ is defined as $\phi_{Ln}^{Ln} = A_{rad} /\left( {A_{rad} + A_{nrad} } \right)$.

$$\tau = \frac{{\sum A_{i} \tau_{i}^{2} }}{{\sum A_{i} \tau_{i} }}$$

(6)

For all doping concentrations, the Pechini-derived phosphors feature the highest $\phi_{Ln}^{Ln}$, confirming that the Pechini synthesis is the best route to get phosphors with better luminescent features. In the Pechini-derived series, $\phi_{Ln}^{Ln}$ increases from 1 to 3% of doping and then, it decreases, while for the phosphors obtained by the sol–gel route, $\phi_{Ln}^{Ln}$ enhances from 1 to 5%, and then it decreases, suggesting that the optimal doping concentration is around 3–5%, which is in accordance with a previous study reported for BaAl₂O₃:Eu³⁺ by R. Chatterjee and coworkers [15], implying that a quenching mechanism by concentration takes place from that on. This mechanism occurs because as the Eu³⁺ doping concentration increases, there occurs a shortening of the Eu–Eu distance within the BaAl₂O₄ lattice, increasing the energy transfer between them. However, this energy transfer may find network defects into the lattice, favoring non-radiative processes, enhancing the non-radiative decay probabilities, and as a consequence, decreasing $\phi_{Ln}^{Ln}$.

Usually, this energy transfer among Ln³⁺ ions occurs non-radiatively by exchange or multipolar mechanisms [4]. The exchange mechanism contribution gains importance as the Ln³⁺-Ln³⁺ distance decreases below 5 Å, since this mechanism occurs by electron transfer and requires a wide orbital overlap. Thus, the critical Ln³⁺-Ln³⁺ distance within the matrix is estimated by means of Eq. 7, where n_A is equal to (x.n)/V, and V is the unit cell volume (V = 830 ų, from the JCPD-17–0306 card) [15], x is the dopant concentration and n is the Ba²⁺ mol number per unit cell (n = 6 from the JCPD-17–0306 card) [15]. As the concentration increases, the Eu³⁺-Eu³⁺ distance becomes shorter, suggesting that the exchange mechanism becomes relevant at higher doping concentrations.

$$R_{C} = \left( {\frac{3}{{4\pi n_{A} }}} \right)^{1/3}$$

(7)

To evaluate if the multipolar mechanism contributes to the energy transfer, the theoretical model developed by Van Uitert [51] was applied which is showed in Eq. 8, where x is the Eu³⁺ concentration, I is the emission intensity (A_rad was considered in this case), and β is a constant. In Eq. 8, θ is equal to 6 for dipole–dipole (D–D) interactions, 8 for electric dipole–quadrupole interactions, and 10 for electric quadrupole–quadrupole (Q–Q) interactions. In this sense, in a graphic of log(I/x) versus log(x), the slope of the curve is equal to –θ/3. For both Pechini and sol–gel-derived samples, θ is 3 (Figure S5, supplementary material), suggesting that the multipolar mechanism is not relevant for the Eu³⁺-Eu³⁺ energy transfer. Thus, the exchange mechanism is dominant on the Eu³⁺-Eu³⁺ energy transfer in the samples featuring high Eu³⁺ concentration.

$$\frac{I}{x} = \left[ {1 + \beta \left( x \right)^{\theta /3} } \right]^{ - 1}$$

(8)

By using $\phi_{Ln}^{Ln}$ as a figure of merit, Table 4, the phosphor efficiency is comparable with other typical red-emitting phosphors featuring high intrinsic emission quantum yield. Moreover, the value reported by us is also larger than the only value reported for BaAl₂O₄:Eu³⁺ by R. J. Wiglusz and coworkers, [11] which confirm that the Pechini route and the charge compensation mechanism are desirable strategies toward enhancing the luminescent features.

Table 4 Comparison of BaAl₂O₄:Eu³⁺ with other red-emitting phosphors by using the intrinsic emission quantum yield ($\phi_{Ln}^{Ln}$) as a figure of merit

Finally, from the structural point of view, both sol–gel and Pechini methods enable the formation of highly crystalline and nanostructured BaAl₂O₄:Eu³⁺ red-emitting phosphor, although the sol–gel process also leads to the formation of traces of the BaCO₃ spurious phase. Moreover, the Pechini route is better to optimize the phosphor synthesis, leading to the highest intrinsic emission quantum yield (86% for PC3) and high emission color purity, which qualifies it for application as a UV-to-red downshifting converter phosphor.

Conclusions

Herein, UV-to-red downshifting converter phosphors based on BaAl₂O₄:Eu³⁺ were synthesized by modified sol–gel or Pechini methods, performing a detailed investigation of luminescence, particle shape and structure. Both methods lead to highly crystalline samples and Eu³⁺ replaces Ba²⁺ within the BaAl₂O₄ lattice; yet, the sol–gel-derived samples display traces of BaCO₃ spurious phase. Moreover, under UV excitation, all phosphors feature emission with high color purity within the red spectral region as a result of electronic transitions from the Eu^{3+ 5}D₀ state to the ⁷F_0-4 ones. Finally, the Pechini-derived sample doped with 3% of Eu³⁺ displays the largest intrinsic emission quantum yield (86%) reported so far for BaAl₂O₄:Eu³⁺, qualifying it for application as a red-emitting phosphor.