99.99% Eu₂O₃ and 99.98% Gd₂O₃ were purchased from Jiangxi South Rare Earth Metals Institute in China. 2-thienyltrifluoroacetonate (HTTA), terephthalic acid (TPA), Phenanthroline (Phen) and other reagents were all analytical grade and used without further purification.

C, H, N analysis was performed on an American Perkin-Elmer 2400 II CHNSLO elemental analyzer. The percentages of rare earth ions were determined by complexometric titration with EDTA. The infrared spectra were measured at room temperature on Nicolet-550 spectrophotometer (American Perkin-Elmer) using KBr pellets in the spectral range of 4000 - 400 cm⁻¹. Differential thermoanalysis (DTA) was performed in a SHDT- 40 thermoanalyticmeter using aluminum crucibles with ~18.40 mg of sample, under dynamic synthetic air atmosphere (40 mL∙min⁻¹) and heating rate of 10˚C∙min⁻¹ in the temperature range of 30˚C - 600˚C. The thermogravimetric (TG) curves were recorded with a thermobalance model SHDT- 40 in the temperature interval of 30˚C - 600˚C, using platinum aluminum with ~18.0 mg of sample, under dynamic synthetic air atmosphere (40 mL∙min⁻¹) and heating rate of 10˚C∙min⁻¹. A Fluorolog FL3-L₁ (American JOBIN YVON) Spectrometer was used to record excitation and emission spectra of the complex powders and their doped silica gels. The bandwidth of monochromators was set at 2.5 nm for both excitation and emission.

The complex Eu₂(TPA)(TTA)₄Phen₂ was prepared in the following steps. In the first step, standard solution of europium (III) (1.0 × 10⁻¹ mol∙L⁻¹) was prepared by dissolving Eu₂O ₃ in hot hydrochloric acid, evaporating up to syrup and diluting with ethanol to a desired volume. HTTA, TPA and Phen were dissolved separately in ethanol with molar ratio of 4:1:2. Subsequently EuCl₃ and HTTA solutions were mixed with molar ratio of 1:2, adjusting pH vales to 5.0, stirred and refluxed for 40 min keeping temperature in water-bath. Then according to molar composition of formula Eu₂(TPA)(TTA)₄Phen₂, Phen and TPA solutions were added dropwise, keeping pH vales 6.5, stirred and refluxed until the appearance of an orange precipitate. The solid product was filtered, washed and recrystallized in ethanol.

The complexes Eu_2(1−x)Gd_2x(TPA)(TTA)₄Phen₂ were prepared by the similar process as Eu₂(TPA)(TTA)₄Phen₂, except that the mixture solution of EuCl₃ and GdCl₃ was used instead of the EuCl₃. The products obtained are an orange precipitate.

Silica gels were prepared by hydrolysis and condensation of tetraethoxysilane Si(OC₂H₅)₄ (TEOS). TEOS and H₂O were used according to molar ratio of 1:4.3 with the proper amount of dimethylformamide (DMF) as solvent. The DMF solution of the complexes was subsequently added to the silica precursor solution (n_complex: n_TEOS = 1:20). Then the mixed solution was refluxed in a 70˚C water bath until gelation occurred. The resulting gels were dried at 100˚C for two weeks for measurement purposes.

The rare earth percentages were determined by complexometric titration with EDTA. Analytical data of C, H, N and rare earth ions percentages (found/calculated) for the complex Eu₂(TPA)(TTA)₄Phen₂ are Eu (III): 17.15/ 17.74; C: 44.76/44.84; H: 2.02/2.10; N: 3.58/3.27; for the complexes Eu_1.4Gd_0.6(TPA)(TTA)₄Phen₂ are rare earth (III): 17.76/17.90; C: 44.60/44.76; H: 2.05/2.10; N: 3.39/3.27; for the complexes Eu_1.0Gd_1.0(TPA)(TTA)₄Phen₂ are rare earth (III): 17.81/18.00; C: 44.57/44.70; H: 1.99/2.10; N: 3.45/3.26; and for the complexes Eu_0.8Gd_1.2(TPA)(TTA)₄Phen₂ are rare earth (III): 18.01/18.05; C: 44.50/44.67; H: 1.97/2.10; N: 3.37/3.26. The elemental analytical data are consistent with the calculated values of the general formula of the Eu (III) complexes.

Table 1 shows some results of IR spectra. The presence of carboxylate groups in the Eu (III) complexes was definitely confirmed by both the asymmetric stretching bands at 1552 - 1558 cm⁻¹ and the symmetric stretching at 1397 - 1401 cm⁻¹. The separations (∆ = υ_as − υ_s) between υ_as (coo) peaks and υ_s (coo) pears are in the range of 155 - 159 cm⁻¹ in the Eu (III) complexes, which are attributed to the bidentate chelating, bidentate bridging and tridentate chelating-bridging coordination modes of carboxylate groups with rare earth ions, since the separations (∆ = υ_as − υ_s) in the Eu (III) complexes are lower than that in Na₂TPA (Δ = 168 cm⁻¹) [19] [20] . Furthermore, owing to the great steric hindrance of the complexes Eu_2(1−x)Gd_2x(TPA)(TTA)₄Phen₂, the bidentate bridging and tridentate chelating-bridging coordination of carboxylate groups with rare earth ions becomes more difficult than the bidentate chelating coordination does. Thus, the coordination mode of carboxylate groups of

TPA with rare earth ions is mainly the bidentate chelating coordination mode in the complexes, and the proposed chemical structure of the complexes is binuclear structure. The IR spectra also show a displacement of υ (C=O) stretching from ~1680 cm⁻¹, in free TTA ligand, to ~1605 cm⁻¹ in the complexes, and a displacement of υ (C=N) stretching from ~1596 cm⁻¹, in free Phen ligand, to ~1559 cm⁻¹ in the complexes, indicating that rare earth ions are coordinated by the oxygen or nitrogen atoms [21] .

The DTA and TG curves of Eu₂(TPA)(TTA)₄Phen₂ and Eu_1.0Gd_1.0(TPA)(TTA)₄Phen₂ are shown in Figure 1 in the temperature range from 50˚C to 600˚C. It can be seen that the DTA and TG curves of the complex Eu₂(TPA)(TTA)₄Phen₂ (Figure 1(A)) present similar profile to that of the complexes Eu_1.0Gd_1.0(TPA)(TTA)₄Phen₂ (Figure 1(B)). Both Eu₂(TPA)(TTA)₄Phen₂ and Eu_1.0Gd_1.0(TPA)(TTA)₄Phen₂ possess good thermal stability, which melt at ~241˚C and decompose at ~370˚C - 430˚C, with no decomposition before the melting point. These indicate that both have similar chemical structure corresponding to the formation of the new complexes. Moreover, in the interval 50˚C - 200˚C, the TG curves of the complexes do not present any event relative to water loss, which indicates that the new complexes are in anhydrous form. This is corroborated by elemental analysis.

The fluorescence excitation and emission spectra of the solid state complexes were performed in a Fluorolog FL3-L₁ Spectrometer.

Figure 2 shows the excitation spectra of the Eu₂(TPA)(TTA)₄Phen₂, Eu_1.4Gd_0.6(TPA)(TTA)₄Phen₂ and Eu_0.8Gd_1.2(TPA)(TTA)₄Phen₂ complex powders recorded in the spectral range of 220 - 450 nm by monitoring the emission at the hypersensitive ⁵D₀ → ⁷ F ₂ transition. These complex excitation spectra show a strong broad band ranging from 250 to 425 nm with the optimum excitation wavelength at ~382 nm, and they are entirely different from the excitation spectrum of Eu³⁺ (without the organic ligands) [22] . These indicate that the ligands have transferred the energy absorbed to the Eu³⁺ ion, leading to the fluorescence enhancement of Eu³⁺.

The fluorescence emission spectra of the complexes Eu_2(1−x)Gd_2x(TPA)(TTA)₄Phen₂ (x = 0 - 1) are all similar except the relative intensity. Figure 3 shows only the emission spectra of Eu₂(TPA)(TTA)₄Phen₂, Eu_1.4Gd_0.6(TPA)(TTA)₄Phen₂ and Eu_0.8Gd_1.2(TPA)(TTA)₄Phen₂ recorded in the range of 560-710 nm with the optimum excitation wavelength at room temperature. It can be seen that five typical Eu (III) emission bands appear at ~582, ~593, ~615, ~654, ~704 nm, corresponding to ⁵D₀ → ⁷F ₀, ⁵D₀ → ⁷F ₁, ⁵D₀→ ⁷F ₂, ⁵D₀ → ⁷F ₃, and ⁵D₀ → ⁷F ₄, respectively. The characteristic emission peaks of Eu³⁺ ions do not change with the addition of co- fluorescence Gd³⁺ ions. Compared with the powdered complexes, the silica gels doped with these complexes show weaker split lines of the Eu³⁺ ion. The phenomenon can be accounted for in the following way. For the powdered complexes, Eu³⁺ ions with surrounding environment of different ligand groups is in a site without a center of inversion, so the emission peaks split into more lines under the ligand field. However, silica gel is a kind of noncrystalline substance with a porous macrostructure. And the complex molecules fixed in the pores are highly ordered. The weaker split lines of Eu³⁺ are observed because the symmetry of Eu³⁺ ion in the powder is lower than that in silica gel.

In Figure 3(A), the relative emission intensity of the complex powder of Eu₂(TPA)(TTA)₄Phen₂ is weaker than that of Eu_0.8Gd_1.2(TPA)(TTA)₄Phen₂. While for the silica gels doped with these complexes, in Figure 3(B), the reverse is true. But compared with the silica gels doped with Eu_1.4Gd_0.6(TPA)(TTA)₄Phen₂, the relative emission intensity of the silica gels with Eu₂(TPA)(TTA)₄Phen₂ remain weaker. The intermolecular energy transfer cannot explain perfectly these changes of the fluorescence intensity.

To investigate the co-fluorescence effect of Gd³⁺ for the complex powders and their doped silica gels, We would assume that co-fluorescence Gd³⁺ ions have no influence on the fluorescence intensity of Eu³⁺ in the complexes, the relative emission intensity value I_calc of the complex powders and their doped silica gels was calculated according to the molar fraction of Eu³⁺ in different co-fluorescence complexes, and the I_exp is the relative emission intensity value for experiment, so the ratios of I_exp and I_calc can be gotten. The results are also listed in Table 2.

As is well known that the energy transfer for the complexes from the ligand to the lanthanide ions can take place via intra molecular energy transfer mode, and the energy transfer can also take place via intermolecular transfer mode. In this study, we think that the co-fluorescence distinction of Gd³⁺ ions for complex powders and their doped silica gels can be preferably interpreted from the proposed binuclear structure together with monomolecular compositions of the complexes. Since the coordination mode of carboxylate groups of TPA with rare earth ions is mainly the bidentate chelating coordination mode in the complexes. The proposed binuclear structure may be given in Schemes 1(a)-(c). For a certain x value in Eu_2(1−x)Gd_2x(TPA)(TTA)₄Phen₂, the complex powders were prepared by coprecipitation, so the complexes are composed of monomolecular EuGd(TPA)(TTA)₄Phen₂, Eu₂(TPA)(TTA)₄Phen₂ and Gd₂(TPA)(TTA)₄Phen₂. In addition, the chemical structure of the mononuclear complex Eu(TTA)₃Phen is also given in Scheme 1(d) [8] .

Figure 4 shows the relationship between the percentages of every monomolecular compositions and the content of Gd³⁺ ions in the complexes. It can be seen from Figure 4 firstly, that with the increase of x value, i.e., the contents of Gd³⁺ ions, the percentages of monomolecular EuGd(TPA)(TTA)₄Phen₂ and Gd₂(TPA)(TTA)₄Phen₂ increase. Since energy can only be transferred from a molecule to other molecules at short distances, and the complex powders were prepared by coprecipitation, the short distance between Gd₂(TPA)(TTA)₄Phen₂ molecules and Eu₂(TPA)(TTA)₄Phen₂ molecules in the coprecipitate makes the intermolecular effective energy transfer be possible. In addition, intra molecular energy transfer in cross binuclear monomolecular EuGd(TPA)(TTA)₄Phen₂ increases. They result in the increase of the emission intensity of the complex powders. When x value is 0.5, the percentage of monomolecular EuGd(TPA)(TTA)₄Phen₂ reaches maximum, the intra

molecular energy transfer reaches also maximum. Then with a further increase of x values the intra molecular energy transfer decreases, on the other hand, intermolecular energy transfer is thought to be concerned with the molar ratios of Eu₂(TPA)(TTA)₄Phen₂ and Gd₂(TPA)(TTA)₄Phen₂. So the optimum concentration of Gd³⁺ in the complex powders Eu_2(1−x)Gd_2x(TPA)(TTA)₄Phen₂ is ~0.5 (molar fraction).

In the silica gel doped with these complexes, the complex molecules are trapped in the pores and isolated from each other. The long distance between complex molecules in the silica gel makes the intermolecular energy transfer impossible. Yet intra molecular energy transfer in cross binuclear monomolecular EuGd(TPA)(TTA)₄Phen₂ is not affect by silica gel. So with the increase of the contents of Gd³⁺ ions, the percentage of the EuGd(TPA)(TTA)₄Phen₂ increases, the fluorescence intensity of the silica gel doped with these complexes increases. But when the contents of Gd³⁺ ions attain to certain value, because of the decrease of Eu₂(TPA)(TTA)₄Phen₂, the fluorescence intensity of the silica gel doped with these complexes decrease. The optimum concentration of Gd³⁺ in silica gel doped is ~0.3 (molar fraction). These interpretations from the binuclear structure together with monomolecular composition of the complexes are consistent with the results of the experiment.