The Structure and Optical Properties of Luminescent Europium Terephthalate Antenna Metal–Organic Frameworks Doped by Yttrium, Gadolinium, and Lanthanum Ions

Abstract

New heterometallic antenna terephthalate MOFs, namely, (Eu_xM_1−x)₂bdc₃·4H₂O (M = Y, La, Gd) (x = 0.001–1), were synthesized by a one-step method from aqueous solutions. The resulting compounds are isomorphic to each other; the crystalline phase corresponds to Ln₂bdc₃∙4H₂O. Upon 300 nm excitation to the singlet excited state of terephthalate ions, all compounds exhibit a bright red emission corresponding to the of ⁵D₀–⁷F_J (J = 0–4) f-f transitions of Eu³⁺ ions. The Eu(III) concentration dependence of the photophysical properties was carefully studied. We revealed that Gd-doping results in photoluminescence enhancement due to the heavy atom effect. To quantitatively compare the antenna effect among different compounds, we proposed the new approach, where the quantum yield of the ⁵D₀ formation is used to characterize the efficiency of energy transfer from the ligand antenna to the Eu³⁺ emitter.

1. Introduction

Metal–organic frameworks (MOFs) represent a large class of crystalline materials defined as porous networks, consisting of metallic ions or clusters linked together by organic multidentate ligands. Due to their well-defined crystallinity, porosity, high stability, and wide diversity of structures and topologies, these materials have attracted considerable attention in the past two decades. Rare-earth element (REE) metal–organic frameworks (MOFs) are of particular interest due to their unique luminescence properties significantly determined by the type of lanthanide ion. Thus, REE-MOFs have been revealed as promising candidates for light-emitting materials, sensors, multimodal image contrast agents, catalysts, and analytes to reveal the hazardous substances in food and enviroment [1,2,3,4,5,6,7]. The f-f transitions are forbidden by selection rules, which results in inefficiency of direct excitation of lanthanide ions. This problem can be overcome by using the energy transfer from the excited ligand to the lanthanide ion, which is called the “antenna effect” [8,9]. The organic ligands, which are used as “antenna” compounds, have high UV absorption coefficients, easily coordinate with REE ions, and efficiently transfer energy to REE ions. The mechanism of the antenna effect can be explained as follows. Upon photon absorption, the ligand in the ground singlet state (S₀) is promoted to the singlet excited state (S₁, S₂, etc.), followed by fast internal conversion to a lower excited energy level (S₁). The excited singlet state can then either (i) return to the ground state (S₁→S₀) by internal conversion and fluorescence or (ii) undergo an intersystem crossing to the triplet state. Through internal conversion, the ligand reaches the lowest triplet level electronic state, T₁, followed by the energy transfer to the REE ion [10]. In Eu-based antenna complexes, the quantum yield of luminescence from the europium ion depends on the relative energy level of the ligand triplet state and the atomic level of the Eu³⁺ ion. Energy transfer has been found to occur from the lowest energy triplet state of the ligand T₁ to the ⁵D_J (J = 0–3) level of the Eu³⁺ ion followed by internal conversion to ⁵D₀. The radiative transitions ⁵D₀–⁷F_J (J = 0–4) to the term of ground state ⁷F of the Eu³⁺ ion correspond to the resulting photoluminescence of such antenna complexes. For the most efficient energy transfer, the difference in energy between the ligand triplet state and the ⁵D_J level of Eu³⁺ should be approximately 2500–4000 cm⁻¹ [11]. However, it has also been proposed that energy transfer can occur from a singlet excited state as well [12,13]. For example, in a study by Shinji Miyazaki et al., the possibility of two energy transfer pathways was revealed from both the triplet and singlet levels for the Eu(hfa)₃(DPPTO)₂ complex (hfa—hexafluoroacetylacetonate, DPPTO—2-diphenylphosphoryltriphenylene) [14].

The simultaneous presence of both luminescent and nonluminescent REE ions, such as Sc³⁺, Y³⁺, La³⁺, Gd³⁺, Lu³⁺, can significantly affect the structural and photophysical properties of these compounds. The structural properties of heterometallic REE-MOFs have been investigated in several studies. It has been found that at low concentrations of the luminescent lanthanide ions, substitution occurs isomorphically, without changing the crystalline structure. However, it has been observed that in some compounds, the structure changes as concentration increases. For example, Jarley Nascimento et al. showed that compounds Gd_1−xEu_x(1,4-bdc)₃(dmf)₂(H₂O)_n (bdc—benzenedecarboxylate, dmf—dimethylformamide; x = 0.01, 0.03, 0.05, 0.07, 0.09) are isostructural with [Eu₂(1,4-bdc)₃(dmf)₂(H₂O)] at the Eu³⁺ content between 1 and 7 at. % (at. %—the relative atom content of a certain lanthanide to all lanthanide atoms). However, at the Eu³⁺ concentration of 9 at. %, the compound was isostructural with [Eu₂(1,4-bdc)₃(dmf)₂(H₂O)₂] [15]. In our previous research, we found that the compound (Eu_xLu_1−x)₂(1,4-bdc)₃∙nH₂O forms different crystal structures depending on the concentration of the Eu³⁺ ions. At the Eu³⁺ concentration range of 6–100 at. %, the samples were found to be isostructural with Ln₂(1,4-bdc)₃∙4H₂O. However at the Eu³⁺ concentration of 0–2 at. %, the samples were isostructural with the Ln₂bdc₃. At the Eu³⁺ concentration between 3 and 5 at. %, both the Ln₂bdc₃∙4H₂O and the Ln₂(1,4-bdc)₃ phases were observed in the samples [16]. The changes in the crystal structure can significantly affect the optical properties of compounds, such as the fine structure of emission spectra, quantum yields, and lifetime values.

Meanwhile, the optical properties of the heterometallic MOFs also depend on the concentration of the luminescent ion in the case of isomorphic substitution of the REE ion by the luminescent ion in the whole concertation range where the single crystalline phase is formed. However, very few works have studied such a concentration dependence. Thus, Utochnikova et al. studied the optical properties of heterometallic solid solutions of (Tb_xY_1−x)₂(1,4-bdc)₃(H₂O)₄ and Eu_xGd_1−x(dbm)₃(phen) (dbm—dibenzoylmethanate, phen—o-phenantroline) MOFs. The steep quantum yield rise was observed at the low concentrations of Eu³⁺ or Tb³⁺ ions. Then, in the range of 20–100 at. % of Tb³⁺ for (Tb_xY_1−x)₂(1,4-bdc)₃(H₂O)₄ and 10–100 at. % of Eu³⁺ for Eu_xGd_1−x(dbm)₃(phen), the quantum yield does not depend on the concentration of the luminescent ions. It has also been observed that lifetimes decrease with increasing concentration of the luminescent ion [17].

We note that the majority of studies focus on compounds with a single or a few concentrations of the luminescent lanthanide ion [17,18,19] or a narrow range of concentrations [15]. The properties of the synthesized compounds have been studied incompletely, and, therefore, at the moment, we have limited information about the mechanism of the dopant concentration effect on photophysical properties of heterometallic REE-MOFs.

In this article, the photophysical properties (photoluminescence decay time constants, radiative, nonradiative, and total decay rates, quantum efficiencies, and formation quantum yields of the ⁵D₀ level; photoluminescence quantum yields, asymmetric ratios) of REE-MOFs of solid solutions of heterometallic (Eu_xM_1−x)₂(1,4-bdc)₃∙4H₂O MOFs (M = Y, La, Gd) were studied in detail in a wide concentration range of the Eu³⁺ ion (0.1–100 at. %).

2. Results

2.1. Structure and Morphology

The phase composition of (Eu_xM_1−x)₂(1,4-bdc)₃·nH₂O (M = Y, Gd, La) with a Eu³⁺ concentration from 0 to 100 at. % was studied using powder X-ray diffraction (PXRD). The experimental PXRD patterns at. %of the synthesized materials are presented in Figure 1a and Figure S1 (Supplementary Materials). The positions of the diffraction maxima in the PXRD patterns indicate that all the synthesized compounds correspond to the Ln₂(1,4-bdc)₃·4H₂O crystalline phase (Ln = Ce − Yb) [20], and no additional peaks were observed. In the Ln₂(1,4-bdc)₃·4H₂O structure (Figure 1b), the octacoordinated lanthanide ions are bound to the two water molecules and six different terephthalate ions through the oxygen atoms.

The refinement of the unit cell parameters and the calculation of the unit cell volumes were performed for the selected samples over the Eu³⁺ concentration range between 0 and 100 at. %. Unit cell parameters (Table S1, Supplementary Materials) were refined using UnitCell software [21]. This program can retrieve unit cell parameters from diffraction data using a least-squares method from the positions of the indexed diffraction maxima of the PXRD patterns (Pawley method [22]). The Eu³⁺ concentration effect on the unit cell volumes is shown in Figure 2. For the (Eu_xLa_1−x)₂(1,4-bdc)₃·4H₂O compounds, the increase in La³⁺ content leads to increased unit cell volumes due to a higher ionic radius of La³⁺ ions (1.160 Å, the coordination number is eight) than the ionic radius of Eu³⁺ ions (1.066 Å) [23].

The ionic radius of the Gd³⁺ ion (1.053 Å) is close to that of Eu³⁺. Therefore, the unit cell parameters do not change significantly in the (Eu_xGd_1−x)₂(1,4-bdc)₃·4H₂O series. The Y³⁺ ion (0.977 Å) is less than Eu³⁺; therefore, substitution of Eu³⁺ by the Y³⁺ ion results in a decrease in the unit cell volumes in the (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O series. In general, the dependence of unit cell volume obeys Vegard’s law [24], with a slight deviation from linearity that falls within the error limits.

Scanning electron microscopy (SEM) was used to reveal the particle morphology and the porosity of the selected synthesized materials, namely, (Eu_0.5M_0.5)₂(1,4-bdc)₃·4H₂O (M = Y, La, Gd). The resulting compounds had a distinct porous structure, as can be seen from the SEM images in Figure 3. On average, the particles are between 5 and 20 µm in size, and the pore diameter ranges from 20 to 150 nanometers. This observation confirms that the synthesized compounds are MOFs according to the definition of IUPAC [25]. Heterometallic europium terephthalates doped with yttrium and gadolinium formed spindle-shaped particles, while those with lanthanum formed flake-like particles.

2.2. IR Spectroscopy

To reveal the doping effect on the vibrational structure of the ligands, we measured the IR spectra of the selected samples of heterometallic (Eu_0.5M_0.5)₂(1,4-bdc)₃·4H₂O (M = Y, Gd, La) and homometallic M₂(1,4-bdc)₃·4H₂O (M = Y, Gd, La, Eu) terephthalates (Figure 4). The broad band with the maximum at about the 3500 cm⁻¹ region corresponds to the O-H stretching vibrations of the water molecules coordinated to the metal. The multiple narrow bands in the 1270–1470 and 1470–1800 cm⁻¹ regions correspond to the symmetric and asymmetric stretching vibrations of the carboxylic -COO group of the terephthalate ion, respectively. The data obtained are consistent with the data in the literature obtained for REE terephthalates [26,27]. The spectral shape, including the fine structure of the absorption bands and the position of the absorption maxima, is almost identical for the studied compounds, which indicates the similar structure of these terephthalates. This conclusion is in agreement with PXRD data, demonstrating that all the synthesized compounds have the same crystalline phase, namely, Ln₂(1,4-bdc)₃·4H₂O.

2.3. Thermogravimetric Analysis

Thermogravimetric analysis (TGA) was performed for selected heterometallic (Eu_0.5M_0.5)₂(1,4-bdc)₃·4H₂O (M = Y, Gd, La) and homometallic M₂(1,4-bdc)₃·4H₂O (M = Y, Gd, La, Eu) terephthalates in the temperature range of 35–200 °C (Figure 5). Weight loss between 8.5 and 9.3% was observed at a temperature of 120–180 °C for all measured samples. As previously reported [28], weight loss in this temperature range may be associated with the dehydration of compounds, leading to the formation of anhydrous terephthalates with a general formula of M₂(1,4-bdc)₃. The weight loss of 8.5–9.3% corresponds to the elimination of 3.8–4.1 water molecules from the initial terephthalates, which is in agreement with the PXRD data, showing that all studied materials are formed in the Ln₂(1,4-bdc)₃∙4H₂O crystalline phase.

2.4. Luminescent Properties

For all synthesized compounds, emission spectra were measured upon 300 nm excitation into the S_n singlet state of the terephthalate ion. Figure 6 shows the normalized emission spectra of selected samples with different Eu³⁺ content ((Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O (M = Y, La, Gd) (x = 0.001, 0.01, 0.1, 0.5, 1)). The emission spectra of all studied materials are given in Figure S2 (Supplementary Materials). All emission spectra contain the same narrow bands, which correspond to the ⁵D₀–⁷F_J (J = 1, 2, 4) transitions of Eu³⁺: ⁵D₀–⁷F₁ (587.9 and 591.6 nm), ⁵D₀–⁷F₂ (614 nm), and ⁵D₀–⁷F₄ (696 nm).

⁵D₀–⁷F₃ transitions are also present in emission spectra, but they were not observed due to their weak intensity. The presence of Eu³⁺ f-f bands in the emission spectra upon excitation to the S_n singlet state of the terephthalate ion clearly reveals an antenna effect. Thus, the terephthalate ion absorbs UV radiation followed by efficient energy transfer to the luminescent lanthanide ion. Upon the excitation, the terephthalate ion is promoted into the S_n state, followed by the fast internal conversion to the S₁ state. Due to the heavy atom effect caused by the lanthanide atom, the S₁ state efficiently undergoes intersystem crossing to the T₁ triplet electronic excited state. The T₁ state of the terephthalate ion is close in energy to the ⁵D₁ energy level of the Eu³⁺ ion [11,17]. Therefore, an efficient energy transfer occurs from the sensitizer to the luminescent lanthanide ion. The ⁵D₁ level of Eu³⁺ then undergoes an internal conversion to the ⁵D₀ state, followed by the emission to the ⁷F_J (J = 1, 2, 4) lower-lying energy levels.

The shape of the emission spectra for the Y-, Gd-, and La-doped compounds is identical to that of the pure europium terephthalate emission spectrum (Figure S3, Supplementary Materials), implying the same coordination environment of Eu³⁺ in the solid solutions studied. This observation agrees with the PXRD data, which show the presence of the same crystalline structure, (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O (M = Y, Gd, La), among the studied series. However, the peak intensity of emission spectra depends on the concentrations of europium ions in the studied solid solutions due to different photoluminescence quantum yields, which will be discussed further.

The photoluminescence decay curves of the (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O phosphors monitored at 615 nm (⁵D₀–⁷F₂ transition) are presented in Figure 7 (λ_ex. = 300 nm). The decay curves were fitted by a single exponential function:

$$I=I_0·e^{{\displaystyle \frac{-t}{\tau }}}$$

(1)

where τ is the observed ⁵D₀ lifetime (

Table 1. The observed ⁵D₀ lifetime of (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O (M = Gd, La, Y; x = 0.01, 0.1).

EuxM1−x	τ(5D0), ms
Eu0.01Gd0.99	0.48 ± 0.01
Eu0.1Gd0.9	0.46 ± 0.01
Eu0.01La0.99	0.44 ± 0.01
Eu0.1La0.9	0.43 ± 0.01
Eu0.01Y0.99	0.47 ± 0.01
Eu0.1Y0.9	0.45 ± 0.01
Eu	0.44 ± 0.01

).

Figure 7 and Table 1 present the photoluminescence decay curves and lifetime data for the selected concentrations of Eu³⁺ ions in (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O (M = Gd, La, Y) (x = 0.01, 0.1, and 1) among the whole concentration range. As can be seen from Table 1, all the samples have similar lifetimes. This allows us to assume that the ⁵D₀ lifetime of the Eu³⁺ ion almost does not depend on the Eu³⁺ content in the 1–100 at. % concentration range of Eu³⁺.

The measured photoluminescence quantum yields (PLQYs) of the (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O are shown in Figure 8a. It should be noted that the ⁵D₀ lifetimes do not depend significantly on the sample composition and are equal to about 0.45 ms. Meanwhile, the photoluminescence luminescence quantum yields increase with an increase in the Eu³⁺ concentration from 1 to 10 at. % due to the number of luminescence sites increasing. In particular, for the Eu–Y series, there is a rapid increase in the quantum yield up to 10%, after which the PLQY remains within small deviations of values from 10 at. % and reaches a plateau, up to Eu 100%. For the Eu–La series, there is also a rapid increase in the PLQY at low concentrations of Eu up to ~12% and a slight decrease in quantum yield to 10% in the range of Eu concentrations from 20 to 100 at. %. The Eu–Gd series is different from the Eu–Y and Eu–La ones: the PLQY has a maximum at the Eu³⁺ content of 10 at. %. The PLQY of (Eu_0.1Gd_0.9)₂(1,4-bdc)₃·4H₂O is equal to 15%, which is 1.5 times greater than that of (Eu_0.1Y_0.9)₂(1,4-bdc)₃·4H₂O and (Eu_0.1La_0.9)₂(1,4-bdc)₃·4H₂O. Upon further Eu³⁺ content increase, the PLQY of (Eu_xGd_1−x)₂(1,4-bdc)₃·4H₂O decreases, whereas the PLQY of (Eu_xY_1−x)₂(1,4-bdc)₃·4H₂O and (Eu_xLa_1−x)₂(1,4-bdc)₃·4H₂O stays about the same (10%).

Luminescence decay is determined by both radiative and nonradiative processes. Radiative decay rate is determined by dipole transition strength and local-field correction. Nonradiative processes include multiphonon relaxation, quenching on impurities, and cooperative processes such as cross-relaxation and energy migration, the influence of which increases with increasing concentration of europium ions in the solid solution. Radiative and nonradiative decay rates of Eu³⁺-doped phosphors can be calculated from the emission spectrum using the 4f–4f intensity theory [29]. The magnetic dipole ⁵D₀–⁷F₁ transition probability $A_{0–1}=A_{M{D}_0}·n_0^3= 49 {\mathrm{s}}^{-1}$. $A_{M{D}_0}$ is the spontaneous emission probability of the magnetic dipole ⁵D₀–⁷F₁, 14.65 s⁻¹, and $n_0$ is the refractive index, 1.5 [30]. Radiative decay rates $A_{0-J}$ (J = 2, 4) of the ⁵D₀–⁷F_J emission transition can be obtained as follows:

$$A_{0–J}=A_{0–1}·{\displaystyle \frac{{\nu }_{0–1}·I_{0–J}}{{\nu }_{0–J}·I_{0–1}}}$$

(2)

where ν_0–J and I_0–J are, respectively, the frequency and intensity of the corresponding transition ⁵D₀–⁷F_J in the emission spectrum. The radiative decay rate is the sum of all the $A_{0-J}$ values $A_{\mathit{Rad}}=A_{0–1}+A_{0–2}+A_{0–4}$. The nonradiative decay rate can be determined using the observed lifetime and the obtained radiative decay rate: $A_{\mathit{total}}=A_{\mathit{Rad}}+A_{\mathit{Nonrad}}={\displaystyle \frac{1}{{\tau }_f}}$; the quantum efficiency of the ⁵D₀ is

$${\eta }_{{}_{,}{}^{5}D_0}={\displaystyle \frac{A_{\mathit{Rad}}}{A_{\mathit{total}}}}={\displaystyle \frac{{n{(}^5{D}_0{–}^7{F}_J)}_{\mathit{em}.}}{n{(}^5{D}_0)}}$$

(3)

Knowledge of both the PLQY (Figure 8a) and the quantum efficiency of ⁵D₀ (Figure 8b) allows us to calculate the quantum yield of the ⁵D₀ formation. The studied terephthalate solid solutions are antenna complexes, where ⁵D₀ is populated as a result of the energy transfer from the initial excited terephthalate ion (S_n(bdc²⁻)→S₁(bdc²⁻)→T₁(bdc²⁻)→⁵D₀(Eu³⁺)). Therefore, the PLQY can be calculated as the direct product of the quantum efficiency of the ⁵D₀ (${\eta }_{{}_{,}{}^{5}D_0}$) and quantum yield of the ⁵D₀ formation (${\Phi }_{\mathit{form}{.}^5{D}_0}$):

$$\mathrm{PLQY}={\eta }_{{{,}^{5}D}_0}·{\Phi }_{\mathit{form}{.}^5{D}_0}$$

(4)

Therefore, the quantum yield of the ⁵D₀ formation (Figure 8c) was calculated as follows:

$${\Phi }_{\mathit{form}{.}^5{D}_0}={\displaystyle \frac{\mathrm{PLQY}}{{\eta }_{{{,}^{5}D}_0}}}$$

(5)

At Eu³⁺ concentrations up to 10 at. %, the quantum yield of the ⁵D₀ formation increases as a result of the number of luminescence sites increasing. For the Eu–Y series, it remains unchanged up to 100 at. % of Eu³⁺, whereas for the Eu–La and Eu–Gd series, it reaches a maximum at 10 at. % of Eu³⁺ and then decays. The maximum values of quantum yield of the ⁵D₀ formation decrease in a dopant row of Gd–La–Y. The QY of ⁵D₀ formation is close to 100%, which indicates a very-high-efficiency energy transfer from the terephthalate antenna to the light-emitting Eu³⁺ ion. We propose that this observation can be explained by the heavy atom effect, which is more pronounced for the Gd³⁺ ion and less pronounced for the Y³⁺ ion.

The emission spectrum of the Eu³⁺ ion includes a very sensitive forced electric dipole transition ⁵D₀–⁷F₂. At the same time, the ⁵D₀–⁷F₁ magnetic dipole transition is not sensitive to environmental changes. Changes in the environment can be judged by the asymmetry ratio, which is determined by the ratio between the ⁵D₀–⁷F₂ and ⁵D₀–⁷F₁ transition intensities. The higher the asymmetry coefficient, the further away the luminescence center is located from the centrosymmetric geometry [31]. The effect of Eu³⁺ ion concentration on local symmetry in samples (Eu_xM_1−x)₂(1,4-bdc)₃∙4H₂O MOFs (M = Y, La, Gd) is shown in Figure 9. It is seen that an increase in the concentration of the Eu³⁺ ion leads to the growth of the asymmetry ratio due to distortion of the crystal structure near the luminescent atom. Starting from 10 at. % of Eu³⁺, the asymmetry ratio reaches a plateau and does not change within the margin of error. Thus, the data once again confirm the fact that three-charged REE ions replace each other isomorphically in crystalline substances.

The abovementioned results clearly show that the presence of a gadolinium ion leads to the luminescence enhancement in Eu–Gd terephthalates, whereas this effect is almost not pronounced in the Eu–La and Eu–Y series. Thus, (Eu_0.1Gd_0.9)₂(1,4-bdc)₃∙4H₂O demonstrates the maximum PLQY among the Eu–Gd series, 15%. The PLQY value is 4–5% higher than that in the Eu–La and Eu–Y series at the same Eu³⁺ ion content (10 at. %). At the higher Eu³⁺ concentrations, the values of PLQY decrease smoothly for the Eu–Gd series, reaching 10% for Eu₂(1,4-bdc)₃∙4H₂O. For the Eu–La and Eu–Y series, the PLQY stays about the same (9–11%) up to Eu³⁺ concentration of 100 at. %. Therefore, we do not observe Eu³⁺ concentration quenching. The Gd-doping effect is also observed for the calculated values of ⁵D₀ formation quantum yield: for (Eu_0.1Gd_0.9)₂(1,4-bdc)₃∙4H₂O, the QY of ⁵D₀ formation values become close to 100%, which indicates a very-high-efficiency energy transfer from the terephthalate antenna to the light-emitting Eu³⁺ ion. The Gd³⁺ ion has an f⁷ configuration and, therefore, increases the probability (rate) of S₁-T₁ intersystem crossing in the terephthalate ion, which is associated with the heavy atom effect manifested by paramagnetic Gd³⁺ ions. To confirm the proposed mechanism, we measured the emission spectra of gadolinium and yttrium terephthalates upon 300 nm excitation into the ¹ππ band of the terephthalate ion. The prominent increase in the phosphorescence band (510 nm) was observed for Gd₂(1,4-bdc)₃∙4H₂O compared with Y₂(1,4-bdc)₃∙4H₂O, which confirms the increase in the intersystem crossing quantum yield resulting from the increase in the S₁-T₁ nonradiative transition rate (Figure 10) as a result of the presence of the Gd³⁺ ion.

3. Materials and Methods

Europium (III) chloride hexahydrate, yttrium (III) chloride hexahydrate, gadolinium (III) chloride hexahydrate, and lanthanum (III) chloride hexahydrate were purchased from Chemcraft (Kaliningrad, Russia). Benzene-1,4-dicarboxylic (terephtalic, H₂(1,4-bdc)) acid (>98%), sodium hydroxide (>99%), nickel(II) chloride hexahydrate (>99%), EDTA disodium salt (0.1 M aqueous solution), and murexide were purchased from Sigma-Aldrich Chemie GmbH (Taufkirchen, Germany) and used without additional purification.

The 0.2 M solutions of EuCl₃, YCl₃, GdCl₃, and LaCl₃ were prepared and standardized using back complexometric titration with EDTA and nickel chloride solution in the presence of an ammonia buffer (pH ≈ 9). A total of 0.6 mole of sodium hydroxide and 0.3 mole of terephthalic acid were dissolved in distilled water to obtain a 1 L solution of a 0.3 M solution of the disodium terephthalate (Na₂(1,4-bdc)).

Heterometallic terephthalates with a general formula (Eu_xM_1−x)₂(1,4-bdc)₃∙4H₂O (M = Y, Gd, La) were obtained by mixing 0.2 M EuCl₃, 0.2 M MCl₃ (M = Y, La, Gd) with 2 mL of 0.3 M Na₂bdc water solution. EuCl₃ and MCl₃ were taken in stoichiometric ratios. The total volume of 0.2 M EuCl₃ and 0.2 M MCl₃ solutions was equal to 1 mL. White precipitates of the resulting terephthalates were separated from the reaction mixture using centrifugation (4000× g) and washed using deionized water 3 times. The samples of compounds were then dried at 60 °C.

The Eu³⁺/M³⁺ (M = Y, Gd, La) ratios in the heterometallic terephthalates were confirmed using energy-dispersive X-ray spectroscopy (EDX) (EDX spectrometer EDX-800P, Shimadzu, Kyoto, Japan). The Eu/M (M = Y, Gd, La) ratios obtained from EDX were consistent with the expected ratios of Eu³⁺/M³⁺ (M = Y, Gd, La) taken for the synthesis for Eu³⁺ content within 1 at. % accuracy for the most of the samples. The EDX data are provided in the Supplementary Materials in Table S2. X-ray powder diffraction (PXRD) measurements were performed with a D2 Phaser (Bruker, Billerica, MA, USA) X-ray diffractometer using Cu K_α radiation (λ = 1.54056 Å). Thermogravimetry curves were obtained using a TG 209 F1 Libra thermo-microbalance (Netzsch, Selb, Germany). The measurement of FTIR spectra was carried out using the IRAffinity-1 spectrometer (Shimadzu, Kyoto, Japan). To carry out photoluminescence studies, the synthesized samples (20 mg) and potassium bromide (300 mg) were pressed into pellets (diameter 13 mm). The photoluminescence data were obtained with a Fluoromax-4 fluorescence spectrometer (Horiba Jobin Yvon, Kyoto, Japan). Lifetime measurements were performed with the same spectrometer using a pulsed Xe lamp (pulse duration: 3 µs). The absolute values of the photoluminescence quantum yields were recorded using a Fluorolog 3 Quanta-phi device (Horiba Jobin Yvon, Kyoto, Japan). All measurements were performed at 25 °C.

4. Conclusions

In this study, we explored the structure and the optical properties of brightly luminescent heterometallic terephthalate antenna MOFs, namely, (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O (M = Y, La, Gd) (x = 0.001–1), which were obtained by precipitation from the aqueous solutions. The crystalline phase of all synthesized compounds corresponds to the Ln₂(1,4-bdc)₃·4H₂O (Ln = Ce − Yb) [20]. Unit cell parameters of the obtained compounds were refined by Pawley method. The replacement of Eu³⁺ ions by the larger La³⁺ ions results in an increase in unit cell parameters in (Eu_xLa_1−x)₂(1,4-bdc)₃·4H₂O compounds, whereas the substitution of Eu³⁺ by smaller Y³⁺ ion results in a decrease in the unit cell volumes in the (Eu_xM_1−x)₂(1,4-bdc)₃·4H₂O series. Unit cell parameters almost do not change in the (Eu_xGd_1−x)₂(1,4-bdc)₃·4H₂O series due to close values of ionic radii of Eu³⁺ and Gd³⁺. The dependence of unit cell volume obeys Vegard’s law [24], which allows one to consider the studied systems as solid solutions. All compounds demonstrate a pronounced antenna effect. Upon 300 nm excitation into S_n singlet state of the terephthalate ion, obtained compounds demonstrate emission corresponding to the ⁵D₀–⁷F_J (J = 1, 2, 4) transitions of Eu³⁺: ⁵D₀–⁷F₁ (587.9 and 591.6 nm), ⁵D₀–⁷F₂ (614 nm, maximal intensity), and ⁵D₀–⁷F₄ (696 nm). The shape of the emission spectra for the Y-, Gd-, and La-doped compounds is identical to that of pure europium terephthalate emission spectrum due to same local symmetry of Eu³⁺ in the studied solid solutions as a result of crystalline phase isomorphism. The ⁵D₀ excited state lifetime does not depend on the sample composition and is equal to 0.45 ± 0.03 ms. Meanwhile, the peak intensity of emission spectra depends on the concentrations of europium ions in the studied solid solutions due to different photoluminescence quantum yields. Thus, the peak emission intensities and photoluminescence quantum yields increase with an increase in the Eu³⁺ concentration from 1 to 10 at. % due to the number of luminescence sites increasing. At larger Eu³⁺ concentrations, for the Eu–Y and Eu–La series, PLQY and the peak emission intensities remain the same (about 9–11%), whereas in the Eu–Gd series, they reach a maximum of 15% at the Eu³⁺ content of 10 at. % and then slightly decrease, reaching 10% in pure europium(III) terephthalate. The quantum efficiency of the ⁵D₀ state was calculated using the relative intensities of ⁵D₀–⁷F_J (J = 1, 2, 4) transitions and ⁵D₀ excited state lifetimes. To quantitatively compare the antenna effect among different compounds, we propose to use the new approach. Based on the values of PLQYs and quantum efficiency of the ⁵D₀ state, we calculated the quantum yields of the ⁵D₀ formation, which reflect the efficiency of energy transfer from the ligand antenna to the Eu³⁺ emitter. The quantum yields of ⁵D₀ formation are larger for Gd³⁺-doped terephthalates. In the MOFs possessing the highest reported compound values of the PLQY, namely, (Eu_0.1Gd_0.9)₂(1,4-bdc)₃∙4H₂O, the value of quantum yield of ⁵D₀ formation is close to 100%, which indicates a very high efficiency of the energy transfer from the terephthalate antenna to the light-emitting Eu³⁺ ion. The Gd³⁺ ion has the f⁷ configuration and, therefore, the increase in the probability (rate) of S₁-T₁ intersystem crossing in the terephthalate ion is associated with the heavy atom effect of paramagnetic Gd³⁺ ions, resulting in the increase in emission intensity, PLQY, and quantum yields of ⁵D₀ formation.