Synthesis, Structure, Morphology, and Luminescent Properties of Ba₂MgWO₆: Eu³⁺ Double Perovskite Obtained by a Novel Co-Precipitation Method

Abstract

Eu³⁺ doped Ba₂MgWO₆ (BMW) double-perovskite was successfully synthesized for the first time by the co-precipitation method. The synthesis procedure, crystal structure, as well as morphology of obtained samples are presented. Domination of the ⁵D₀–⁷F₁ magnetic–dipole over forced electric–dipole transitions in the emission spectra indicates that Eu³⁺ ions are located in the high symmetry site with inversion center. Only one emission line assigned to the ⁵D₀–⁷F₀ transition was observed, confirming that europium substituted for only one host cation site. The photoluminescence excitation (PLE) spectrum is dominated by a strong and broad band related to the O²⁻ → Eu³⁺ and O²⁻ → W⁶⁺ charge transfer. The decay of the emission from the ⁵D₀ and ⁵D₁ levels was investigated. The temperature-dependent emission spectra showed that the T_0.5 is equal to 350 K. Extinguishing mechanisms of the Eu³⁺ luminescence in the studied host are discussed.

1. Introduction

Recently, Ba₂MgWO₆ (BMW), a perfect cubic crystal structure and very promising representative of double perovskite family, has been put into research in the field of ceramic fabrication by virtue of its excellent dielectric properties [1,2]. However, very few studies have discussed its luminescent properties in detail. Indeed, there are only two studies associated with this host doped with Eu³⁺ ions [3,4]. Still, in both cases, the solid-state method was employed to obtain BMW: Eu³⁺.

The chemical formula of a double perovskite is described as A₂BB’O₆ in which A²⁺ ions are located in 12-coordination sites while B²⁺ and B’ ions are located in 6-coordination sites [5]. The visualization of the Ba₂MgWO₆ crystal structure with all details was published in our previous study [3]. In the BMW structure, Ba²⁺ is located in the site with T_d symmetry. As we showed previously, the trivalent europium ions replace the Mg²⁺ in the site with O_h symmetry with inversion center [3].

In this study, a series of Ba₂MgWO₆: Eu³⁺ was synthesized for the first time by using the co-precipitation method. To exclude defects when the trivalent ion replaces the bivalent one, a Li⁺ co-dopant has been added to locally compensate for the charge. Li⁺ was chosen because of similar ionic radii, which are as follows for ions with six-fold coordination: 108.7, 90, and 86 pm for Eu³⁺, Li⁺, and Mg²⁺, respectively [6]. The charge compensation strategy could be described in this way:

2Mg²⁺ → Eu³⁺ + Li⁺

The purpose of this work is to obtain BMW by wet chemistry methods, i.e., co-precipitation. Of the various synthesis methods, the co-precipitation one is expected to produce material with smaller particle sizes and more homogeneous morphology. A lower sintering temperature contributes to lower energy consumption compared to other synthesis methods [5]. The usefulness of this method will be assessed by comparing the results obtained with the results already published for the solid-state samples. Here, we present the absorbance, excitation, and emission spectra as well as the emission decay profiles. The influence of temperature and dopant concentration on luminescence was also studied.

2. Experimental

2.1. Synthesis

In this study, concentration series of Ba₂MgWO₆: x% Eu³⁺ double perovskite structure (x = 0.1%, 0.5%, 1%, 3%, 4%, 5%, and 7%) were synthesized first time by co-precipitation method, with lower sintering temperature, in comparison with other methods [3,4,7]. To compensate for the evaporation of magnesium ions during the sintering process, a 20% excess of magnesium ions was applied. Ba(CH₃COO)₂ (Alfa Aesar, 99%), Mg(CH₃COO)₂∙4H₂O (Alfa Aesar, 99.95%), (NH₄)₁₀H₂(W₂O₇)₆ (Sigma–Aldrich, 99.99%), Eu(CH₃COO)₃ (Alfa Aesar, 99.9%), were used as starting materials. Firstly, the stoichiometric quantities of barium acetate Ba(CH₃COO)₂, magnesium acetate Mg(CH₃COO)₂∙4H₂O and ammonium paratungstate (NH₄)₁₀H₂(W₂O₇)₆ (APT) were dissolved separately in distilled water. Next, a white precipitate was formed immediately after the first droplet of APT solution added slowly (2 ml/min) under stirring 200 rpm, 25 °C. The precipitate was evaporated by heating at 80 °C for 20 h before pre-sintering at 600 °C for 12 h. The final annealing was carried out at 1150 °C for 6 h, with the constant heating rate, 3 °C/min using corundum crucible. After each step, the obtained products were ground for 15 minutes. The scheme of BMW preparation is presented in Figure 1.

2.2. Characterization

The structure of materials was analyzed by X’Pert ProPANalytical X-ray diffractometer (PANalytical, Almelo, The Netherlands) using Cu K_α radiation (λ = 1.54056 Ǻ) in a 2$\theta$ range from 10° to 90° with a step of 0.026°. For measurements of the reflectance absorption spectra, the Varian Cary 5E UV–Vis-NIR spectrophotometer (Agilent, Santa Clara, CA, USA) was used. The emission spectra in the temperature range 10–300 K were recorded using a Jobin-Yvon monochromator (Horiba Scientific, Kyoto, Japan), along with a closed-cycle helium cryostat. Decay profiles were recorded with a Lecroy digital oscilloscope (Teledyne LeCroy, New York, NY, USA) with the Nd: YAG laser as the excitation source. Scanning electron microscope FEI NOVA NanoSEM230 (FEI, Hillsboro, OR, USA) was used to characterize the morphology of sample. The excitation spectrum was recorded using a McPherson spectrometer with a 150 W xenon lamp as the excitation source and a Hamamatsu R928 photomultiplier (Hamamatsu Photonics K.K, Shizuoka, Japan) as the detector. Temperature-dependent emission spectra were measured with the Hamamatsu Photonic multichannel analyzer PMA-12 equipped with a BT-CCD linear image sensor (Hamamatsu Photonics K.K, Shizuoka, Japan). The temperature of the samples during emission measurements was controlled by the Linkam THMS 600 Heating/Freezing Stage (The McCRONE group, Westmont, IL USA).

3. Results and Discussion

Ba₂MgWO₆ possesses the cubic crystal structure with a rock salt lattice of which corner-shared MgO₆ and WO₆ octahedrons. The Ba cations are located in the 12-fold coordination sites. The crystal structure of BMW was described in detail in our previous study [3]. The X-ray diffractograms of Ba₂MgWO₆: x Eu³⁺ samples (x = 0.1%, 0.5%, 1%, 3%, 4%, 5%, and 7%) are shown in Figure 2. The X-ray diffraction patterns of prepared samples match well with the Ba₂MgWO₆ pattern (ICSD card number 024-982) with the space group Fm-3m. The higher concentration of europium causes the additional diffraction peaks to appear. The alien phases belong to the barium tungstate family, including BaWO₄ (present at 2$\theta$ of 26.4°), Ba₂WO₅ (28°, 28.6°, 29.6°), and Ba₃WO₆ (29.4°, 42°) these phases were also mainly found in other reports [1,7]. Fortunately, the alien phases did not affect the luminescence properties as well as further ceramic preparation. The positions of the XRD peaks were corrected with the aid of Si admixture (ICSD card number 026-1481). Ba²⁺ is a very large ion, its crystal radius (CR) for coordination number equal 12 (CN = 12) is 175 pm. While the CR of the Eu³⁺ ion for CN = 9 is-Shannon gives no data for the highest CN-only 126 pm [6]. If europium ions have replaced barium ions, then the unit cell of BMW should shrink. Since it is growing (see Figure 2c), we conclude that larger Eu³⁺ ions (CR = 108.7 pm) replace smaller Mg²⁺ ions (CR = 86 pm) in the sites with CN = 6. The replacement of W⁶⁺ with Eu³⁺ was excluded because of the large difference between the charge of these ions and their crystal radius equal to 74 and 108.7 pm for W⁶⁺ and Eu³⁺, respectively.

The microstructure of the representative sample BMW: 3% Eu³⁺, and the crystallite size distribution of the obtained sample (calculated using ImageJ software) are shown in Figure 3. In general, single grains were not observed in the SEM image. As can be seen, the morphology of the sample was heterogeneous. However, a large number of crystallites were agglomerated, forming large objects with irregular shapes. As a result, the distribution of the crystallite sizes is in a wide range from 70 nm up to 500 nm. The average crystallite size was estimated to be around 209 nm. This value is much smaller than those obtained using a conventional solid-state technique [3,7], demonstrating one of the advantages of the wet chemistry method. This result encourages further work to eliminate agglomeration to ease ceramic preparation.

The photoluminescence of Ba₂MgWO₆ doped with a series of europium concentration from 0.1% to 7% was measured under 266 nm excitation wavelength at room temperature (see Figure 4). The picture is uncommon to majority cases when Eu³⁺ ions are located at the sites without the center of inversion. Here the ⁵D₀ → ⁷F₁ magnetic dipole transition dominates the spectrum as sharp and intense peak at 596 nm. The forced electric dipole transitions are much weaker and broadened due to vibronic coupling, assigned to the ⁵D₀ → ⁷F_J transitions (where J = 0, 2, 3, 4) are observed at about 584.4, 618.7, 665, and 721.3 nm, respectively. The integrated emission intensity in the function of the Eu³⁺ concentration was depicted in the inset of Figure 4. The luminescent intensity increases significantly with the increase of Eu³⁺ concentration up to 5%. Above that point, the emission intensity decreases as a consequence of the concentration quenching process.

Due to the hypersensitive nature, the ⁵D₀ → ⁷F₂ electric-dipole transition it susceptible to the even small changes of the crystal field. In contrast, the ⁵D₀–⁷F₁ magnetic dipole transition is virtually unaffected by neither symmetry site nor host lattice-type. Hence, the asymmetry factor R, the ratio between the integrated intensity of the ⁵D₀ → ⁷F₂ and that of ⁵D₀ → ⁷F₁ transitions, is often used to determine the changes in the nearest environment of Eu³⁺ in the lattice. Thus, the higher the value R is, the lower the symmetry of the Eu³⁺ ions occupancy site. For the investigated samples, the R parameter increases gradually from 0.56 to 0.63 when dopant concentration increases. It demonstrates a slight distortion of the structure, which insignificantly decreases local symmetry of the Eu³⁺ ions.

To assign every single emission peak, and to determine the number of the Eu³⁺ sites, the emission spectra were recorded at low temperature (77 K and 10 K). The intensity of the f-f lines significantly increased, while the vibronic ones almost disappeared when the temperature decreased down to 10 K (see Figure 5). Only one narrow line at 585 nm attributed to the ⁵D₀ → ⁷F₀ transition was observed at room and low (10 and 77 K) temperature. It is apparent evidence of the europium location in the only one crystallographic site (see the inset in Figure 5). Emission spectra were carefully studied. All detected lines were tentatively assigned either to MD or ED or vibronic transitions. Their energies are presented in Supplementary Table S1.

The excitation spectra of the 1%, 3%, and 5% Eu³⁺ doped Ba₂MgWO₆ was monitored at 596 nm, i.e.at the maximum of the ⁵D₀ → ⁷F₁ transition at room temperature (see Figure 6). The spectra consist of a broad and intense band ranging from 250 nm to 350 nm. The broad band results from two charge transfer transitions (CTB), including the O²⁻ → W⁶⁺ and O²⁻ → Eu³⁺ transitions. There is also a small band at about 395 nm assigned to intra-configurational f-f transitions of Eu³⁺ ions, weak due to the Laporte selection rule [8]. They become partially allowed because of the lowering of the symmetry site when introducing a higher amount of Eu³⁺ and Li⁺ ions. Our recent investigation does not reveal any trace of f-f transitions in the excitation spectra, but there effective doping was several times smaller [3]. The sample with 1% Eu³⁺ virtually does not show the f-f transitions. Mainly CTB is observed similarly as in the solid-state method [3].

It is not easy to find the energy of the O²⁻ → Eu³⁺ transitions since the amount of dopant is small comparing to the tungstate, which forms the host structure. Analogously, the broad band at 318 nm results from both charge transfer transitions in Ba₃WO₆: Eu³⁺ [9]. The energy of the O²⁻ → Eu³⁺ CTB is influenced not only by the ligand electronegativity but also by the local surroundings of the Eu³⁺ ions [8]. The location of the O²⁻–W⁶⁺ CTB has been reported around 300 nm in g-C₃N₄/Ba₂MgWO₆: Eu³⁺ [4], 301 nm in Ba₂MgWO₆: Eu³⁺ [3], 310 nm in host Ba₂MgWO₆ [7], 325 nm [10] in NaLaMgWO₆: Eu³⁺. As noted in the preliminary work, the O²⁻–Eu³⁺ CTB was found at around 304 nm in CaMoO₄: Eu³⁺, 293 nm in AMoO₄: Eu³⁺ (A = Ba, Sr), and at 250 nm in BaNa₂W₂O₁₁: Eu³⁺. The less intense and narrower bands appeared near the visible region (from 370 to 400 nm) and at 462.5 nm that could be ascribed to f-f transitions of the ⁷F₀ → ⁵D₄, ⁵G₇, ⁵L₇, ⁵L₆, ⁵D₃ and ⁷F₀ → ⁵D₂ transitions, respectively, [11,12,13] (see inset in Figure 6).

The absorption spectra registered at 300 K were useful to calculate the energy of the forbidden band-gap of the investigated samples. The band-gap energy E_g was determined by applying the modified Kubelka–Munk function and plotted as (F(R).hv)² versus hv, also called a Tauc method [14]. It was found that the higher dopant content was used, the smaller band-gap energy was obtained, (see Figure 7). For the lowest Eu³⁺ concentration, the band-gap energy was 4.02 eV while for the sample doped with 5% of Eu³⁺, E_g = 3.82 eV.

The 77 K decay curves of the BMW sample doped with 1% Eu³⁺ was excited at 266 nm and monitored at the different wavelengths corresponding to the transitions from the ⁵D₁ (588 nm, 602 nm) and the ⁵D₀ (596 nm, 722 nm) levels (see Figure 8). The decay curves are multi-exponential in all cases. The first component was usually connected with the nonradiative process, while the longer ones resulted from the emissions from different Eu³⁺ energy levels depending on the monitored wavelength. The decay of magnetic-dipole transition consisted of two components, one with the short lifetime-related to the nonradiative process-and the longer one, τ = 5.71 ms, characteristic for the Eu³⁺ located in the site with very high symmetry. The decay time of the ⁵D₁ level was found to be equal to 70 μs, it is much smaller than expected, probably because of the fast nonradiative transition to the lower ⁵D₀ level and also possible cross-relaxation mechanism—the latter will be discussed further.

The luminescence kinetics were recorded at 77 K upon excitation at 266 nm. Figure 9 presents the ⁵D₀ → ⁷F₁ emission decay profiles of the BMW doped with 0.5%, 3%, and 7% of Eu³⁺ ions. The emission decay curves exhibited the characteristics of the nonexponential function. The average decay time can be expressed as follow:

$${\tau }_{avg}=\frac{{{\displaystyle \int }}_0^{\infty }\left(I\left(t\right)\times t\right)dt}{{{\displaystyle \int }}_0^{\infty }I\left(t\right)dt}\cong \frac{{{\displaystyle \int }}_0^{t^{max}}\left(I\left(t\right)\times t\right)dt}{{{\displaystyle \int }}_0^{t^{max}}I\left(t\right)dt}$$

(1)

where I(t) is the emission intensity at time t, 0 < t < t^max. The emission average decay time gradually decreased from 5 ms to 2.9 ms with an increase of Eu³⁺ concentration from 1% to 7%, respectively. Two factors could explain this observation, one is the nonradiative processes, which are undoubtedly present. Second the distortion of the Eu³⁺ coordination polyhedra due to the increase of the dopant concentration. We recall here that the effective doping is twice as important as in the previous study [3,4] because of the presence of Li⁺. The distortion causes a slight decrease in the symmetry, so the transition probability increases, although the emission still dominates the MD transition.

The Judd–Ofelt (J-O) intensity parameters (Ω_λ, where λ = 2, 4, 6) were calculated on the basis of the BMW: Eu³⁺ (5%) emission spectrum excited at 266 nm and recorded at 300 K. Exactly the same formalism and the same set of equations were used as in our previous work [3]. The refractive index “n” of the BMW host is 1.874 [3]. The matrix elements U(2) = 0.0035, U(4) = 0.003 and U(6) = 0.0005 from the work of K. Binnemans [8] were applied for the calculations.

Using the Ω_λ parameters, the A_ij transition rates, the β_ij branching ratios, were calculated for the ⁵D₀ → ⁷F_J transitions (where J = 1, 2, 4, 6) (see

Table 1. The transition rates, branching ratios for ⁵D₀ → ⁷F_J (J = 1, 2, 4, 6) transitions of BMW: Eu³⁺.

Sample		5D0 → 7F1	5D0 → 7F2	5D0 → 7F4	5D0 → 7F6
Co-precipitation	A0-J (s−1)	96.35	77.64	17.82	9.61
	Β0-J (%)	47.8	38.55	8.85	4.8
Solid-state	A0-J (s−1)	96.35	70.33	14.42	7.13
	Β0-J (%)	51.2	37.4	7.6	3.8

and

Table 2. Total, radiative and non-radiative transition rates, calculated and experimental decay time and quantum efficiency of BMW: Eu³⁺.

	Atot (s−1)	Arad (s−1)	Anrad (s−1)	τrad (ms)	τexp (ms)	η (%)
Co-precipitation	223.2	201.4	21.8	4.96	4.48	90.2%
Solid-state	222.2	188.2	34	5.31	4.5	85

). A recent analysis in the frame of J-O theory for the BMW: Eu³⁺ (2%) solid-state sample was carried out at 395 nm of excitation [3]. However, there is a possibility that excitation at 395 nm will be absorbed by all Eu³⁺ ions-including those which could be in the additional, foreign phases present in the sample. Consequently, this time, the emission was excited with the 4^th harmonic of the Nd: YAG laser reaching the O^2- → Eu³⁺ charge transfer band, both for the solid-state and co-precipitation samples.

There is very little difference among J-O parameters between sample synthesized by co-precipitation and solid-state method (see Table 1, Table 2 and

Table 3. The Judd–Ofelt Ω_𝜆 ($\times {10}^{-20}{\mathrm{cm}}^2$) parameters for Eu³⁺ ions in various double perovskites.

	Ω2	Ω4	Ω6	Reference
Ba2MgWO6: 5% Eu3+ (co-precipitation)	1.18	0.48	2.5	This work
Ba2MgWO6: 2% Eu3+ (solid-state)	1.07	0.39	1.83	This work
LiLaMgWO6: 0.01% Eu3+ (solid-state)	7.27	2.59	1.64	[15]
NaLaMgWO6: 0.01% Eu3+ (solid-state)	7.29	1.86	1.8	[15]
KLaMgWO6: 0.01% Eu3+ (solid-state)	7.53	5.32	5.09	[15]

). The value of Ω₂ is twofold smaller than that of Ω₆. The increase in the concentration of Eu³⁺ from 2% to 5% leads to a modification of the Ω_λ (see Table 3). These are phenomenological parameters which only depend on the matrix. An increase in the concentration of dopant ions having a larger diameter and a different charge than the replaced Mg²⁺ must cause the deformation of the matrix. We must not forget either that with 5% Eu³⁺, we introduce 5% Li⁺.

The Ω_𝜆 values found in this work are comparable with other results obtained in similar tungstate double perovskites but with lower symmetry (Li, Na, K)LaMgWO₆ [15]_. Due to the fact that Ω₂ is very sensitive to the angular changes while Ω₄ and Ω₆ are mostly affected by bond covalence between dopants and ligands [16]. As a result, a distinction among various tungstate double perovskites was observed, Ω₂ values of (Li, Na, K)LaMgWO₆ with lower site symmetry are seven times higher in comparison with those of the higher site symmetry (O_h) in BMW. The Ω₄ value of (Li, Na, K)LaMgWO₆ is also fivefold higher than those of BMW while the Ω₆ value is quite similar, except for KLaMgWO₆.

Besides, the rates of total, radiative, non-radiative transition along with the theoretical radiative and experimental decay time as well as the quantum efficiency of sample BMW: Eu³⁺ are presented in Table 2. For sample prepared by co-precipitation and solid-state, the calculated radiative decay times are 4.96 and 5.31 ms and the quantum efficiencies are 90.2% and 85 %, respectively.

The temperature-dependent emission spectra under excitation at 266 nm were performed in the temperature range from 80–730 K (Figure 10a). The emission intensity does not change much up to 200 K. Above this temperature, emission intensity significantly decreases. The temperature quenching, at which the integrated intensity of emission decreased half in comparison with the initial emission intensity (T_0.5) was at around 350 K (Figure 10b). The emission intensity of transitions from the ⁵D₁ level was also quenched, along with the transitions from the ⁵D₀ level. Figure 10 shows the relationship of ln(I₀/I-1) versus 1/kT, two mechanisms responsible for emission quenching took place. The activation energies for thermal quenching were calculated to be 0.09 eV and 0.23 eV. The first one is due to the process which operates in the 80–350 K temperature range. Above 350 K, the second process is switched on and is more effective in the emission quenching. The first process is of the same nature as that observed in LaAlO₃: Eu³⁺ and explained in detail by Blasse [17], but the CT state involved here is associated with the O²⁻ ↔ W⁶⁺ transition. Here, the same approach was applied to support the single configurational coordinate model depicted in Figure 10d. Upon excitation into the charge transfer state-transition A → B, some part of the excitation energy relaxes to point C and then to Eu³⁺, and the rest, with thermal excitation, flows from the CTS to the Eu³⁺ ground state via point D. (see Figure 10d). The other nonradiative process is switched on above 350 K when the ⁵D₁ level becomes thermally populated and is due to the following cross-relaxation: [⁵D₁, ⁷F₀] → [⁵D₀, ⁷F₃] (see points F and E in Figure 10d). Its activation energy (∆E_a = 0.23 eV = 1855 cm⁻¹) is related to the difference energy between the ⁵D₀ and ⁵D₁ and ⁷F₀–⁷F₃ levels.

4. Conclusions

The co-precipitation method was successfully employed in the synthesis of Ba₂MgWO₆ double perovskite doped with Eu³⁺ ions. Dopant replaces with Mg²⁺ and is located at the O_h site with the inversion center. The mean size of the particles determined from SEM images was around 209 nm. The single line of the ⁵D₀–⁷F₀ transition and the dominance of the ⁵D₀–⁷F₁ one are undeniable evidence that europium ions occupy one highly symmetrical site in the BMW host. The strongest emission was observed for 5% of Eu³⁺. The broad band from 250 to 350 nm in the PLE spectrum corresponds to the O²⁻ → Eu³⁺ and O²⁻ → W⁶⁺ charge transfer transitions. Our results are in good agreement with Blasse’s theory, which describes the relationship between Eu³⁺–O²⁻ distance with localization of CTS maximum and emission efficiency. The thermal quenching investigation showed that cross-relaxation processes generally quench the emission of Eu³⁺ at temperatures higher than 350 K. At lower temperatures, the excitation energy is lost through the crossing point of the CTS state and the Eu³⁺ ground level. We believe that this material will be perfect for producing transparent or translucent ceramics.