Crystal structure, thermodynamic behavior, and luminescence properties of a new series of lanthanide halogenated aromatic carboxylic acid complexes

2.1 Materials

As reagent grade materials, the products purchased in the market can be used without further purification of the reaction. All the reagents used in the experiments are listed in supporting information Table S1.

2.2 Experimental equipment and conditions

Mo-Kα and Cu-Kα rays (λ = 0.71073 Å) monochromatized by graphite were used as the incident light source. The diffraction data of single crystals 1–5 were measured at room temperature on a single crystal X-ray diffractometer Smart-1000 (Bruker AXS). After collecting the data, the SHELXS-97 program (Sheldrick and Schneider, 1997) in the computer was used to calculate the analysis. A direct method was used to determine the structures of complexes 1–5, and full-matrix least-squares refinement on F² was used to refine the structures. Measurements were performed in the wavelength range 4000–400 cm⁻¹ using the KBr press method by Bruker TENSOR27 infrared spectrometer. Powder X-ray diffraction experiments were performed on five complex powders by a Bruker D8 ADVANCE X-ray diffractometer at room temperature, using copper-potassium radiation (λ = 0.71073 Å) scanned in the range of 5-50°. The Raman spectra of the complexes were measured using a Bruker VERTEX-90 FTIR-RAMANII in the wavelength range of 4000–50 cm⁻¹ with an Nd:YAG laser (λ = 1.064 μm) and a liquid nitrogen cooling device with a resolution of 4 cm⁻¹ and a laser power of 400 mW and 100 scans in the wavelength range. The TG/DSC-FTIR experiments were performed in a simulated dynamic air atmosphere (20 mL/min of nitrogen, 10 mL/min of oxygen, and 30 mL/min of nitrogen as the protective gas). Experiments were performed using a NETZSCH STA 449 F3 and a BRUKERTENSOR27 Fourier transform infrared spectrometer and equipped with a liquid nitrogen system for the experiments, using an alumina trioxide crucible for release and a blank alumina trioxide crucible as a reference. The heating rate is 10 K/min and heated to 800℃. The fluorescence spectra were measured with an FS5 fluorescence spectrophotometer. The excitation spectra and emission spectra of complexes 3 and 4 were measured under Xe lamp irradiation, and complex 3 fluorescence lifetimes were determined. The Gaussian09 program package was used for the theoretical part of the calculations. Density functional theory (Geerlings et al., 2003) is a quantum mechanical method to study the electronic structure of the multi-electron system. In this case, the B3LYP hybridization generalization (Tirado-Rives and Jorgensen, 2008) is used, using the nonlocal correlations provided by the LYP expressions. The 6-311G(d,p) cleavage valence bond group is chosen, which describes the system wave function better than the STO-3G group.

2.3 Synthesis

The acidic ligand 2,4-DFBA (6 × 10^-4 mol) and the ligand phen (2 × 10^-4 mol) were mixed and dissolved in 6 mL of 95% ethanol and in 3 mL of distilled water, lanthanide nitrate (2 × 10^-4 mol) was dissolved. After adjusting the pH to 5–7 by adding sodium hydroxide solution with a concentration of 1 mol/L to the ligand solution, it was poured into the lanthanide nitrate solution. Stirred at room temperature with a magnetic stirrer for about 6 h and left to stand for about 12 h. Using the room temperature solution volatilization method, the filtrate was extracted and placed in a small clean beaker, covered with a layer of cling film with small holes, and left to stand at room temperature for about a week to obtain the crystals.

Elemental analysis (in percent content): C₆₆H₃₄F₁₂La₂N₄O₁₂, Calcd: H, 2.17; C, 50.15; N, 3.54. Found: H, 2.16; C, 50.35; N, 3.72. C₆₆H₃₄F₁₂Gd₂N₄O₁₂, Calcd: H, 2.12; C, 49.01; N, 3.46. Found: H, 2.07; C, 49.03; N, 3.49. C₆₆H₃₄F₁₂Tb₂N₄O₁₂, Calcd: H, 2.12; C, 48.91; N, 3.46. Found: H, 2.09; C, 48.92; N, 3.43. C₆₆H₃₄F₁₂Dy₂N₄O₁₂, Calcd: H, 2.11; C, 48.69; N, 3.44. Found: H, 2.11; C, 48.98; N, 3.35. C₆₆H₃₄F₁₂Ho₂N₄O₁₂, Calcd: H, 2.10; C, 48.55; N, 3.43. Found: H, 2.06; C, 48.70; N, 3.42.

2.4 Determination of X-ray single-crystal structure and structural analysis

The complexes 1–5 with good crystal quality were selected and measured in a single crystal X-ray diffractometer Smart-1000 (Bruker AXS). The radiation source was Mo-Kα and Cu-Kα rays (λ = 0.71073 Å) monochromatized by graphite, and their crystal diffraction data were measured at room temperature. The rhythmic structures were structurally resolved by the direct method using the SHELXS-97 program (Sheldrick and Schneider, 1997) and refined by full-matrix least-squares on F² to obtain the crystal structures. All atoms in the structure are non-isotropic except for hydrogen. Based on the above data, the crystal.cif file is imported into Mercury/Diamond software to visualize the crystal structure and facilitate the study. The crystal model is drawn in Diamond software according to the structure in Mercury software. As a 3D crystal simulation software (Pennington, 1999), Diamond can display crystal models very well, rotate, move, and scale freely as needed, and make reasonable changes to the coloring scheme. For the coordination environment of the central metal ion, the simulation calculation was carried out by using the shape software. Via the continuous shape measure method (Alvarez et al., 2014, Terebilenko et al., 2022). This software can well classify and organize the topological types and solve the previous errors with the naked eye observation. In complexes 1–5, the 2,4-difluorobenzoate ligand adopts a total of three coordination modes (Zhou et al., 2021) to Ln(III), namely double-dentate chelate (Fig. 1a), bridged double-dentate (Fig. 1b) and bridged tri-dentate (Fig. 1c).

Close

Fig. 1

(a-c) Three bonding modes of 2,4-difluorobenzoate.

Export to PPT

3.1 Description of the crystal structure

The complexes 1–5 were analyzed by single crystal X-ray diffraction. The crystallographic data and the corresponding structural refinement parameters are listed in Table 1. The selected bond lengths and angles are listed in supporting information Table S2, Table S3. Hydrogen bond data are shown in supporting information Table S4. Upon analysis of the data, complexes 1–5 are all monoclinic crystal systems with space group P2₁/n. All compounds are isostructural. It is noteworthy that complexes 1, 2 and complexes 3–5 have the same structural formula, both crystal systems and space point groups are the same. However, they have different crystal structures, coordination environments, and coordination numbers. This may be due to the contraction of the lanthanum system. Therefore, two other crystal structures, represented by complexes 1 and 4, are selected for detailed discussion below.

3.1.2

3.1.2 Structural description of [Ln(2,4-DFBA)₃(phen)]₂, (Ln = La(1), Gd(2))

The crystal structure of 1 and 2 has a muffin-shaped spatial geometry (Fig. 2b), and this configuration was calculated (Madanhire et al., 2020) by the shape software. This configuration is used for complexes 1 and 2 with coordination number nine. Introduce complex 1 as an example in detail. The structural unit of complex 1 (Fig. 2a) consists of two central La(III), six 2,4-difluorobenzoate ligands, and two o-phenanthroline ligands. It has the same two central La(III) coordination environments as a binuclear molecule. Among them, the two nitrogen atoms (N1, N2) that were ligated with the central ion were derived from the o-phenanthroline ligand. The remaining seven oxygen atoms were all derived from the 2,4-difluorobenzoate ligand. Interestingly, these seven oxygen atoms have three coordination modes. They are bridged tri-dentate (O1#, O1, O2), bridged double-dentate (O3, O4), and double-dentate chelate (O5, O6). The bond length of the La-O bond ranges from 2.412(3)-2.801(3) Å. The critical length is similar to that reported in the literature (Shen et al., 2018, Zhu et al., 2018). The average bond length of the La-O bond is significantly longer than that of the other conformation of the complex, which is due to the bond length of La1-O1# involved in the tridentate coordination. As shown in Fig. 3a, through hydrogen bonding along the crystallographic c-axis, complex 1 formed a 1D chain structure. In addition, the adjacent structural units formed a 2D faceted supramolecular structure along the crystallographic ac-plane by C-H···F hydrogen bonding (Fig. 3b) with forces of 3.243 and 3.528 Å, respectively.

Close

Fig. 2

(a) Structural unit of complex 1 and 2 (Ln = La, Gd). (b)The polyhedral coordination environment of central Ln(III) (Ln = La, Gd).

Export to PPT

Close

Fig. 3

(a) The one-dimensional chain structure along the crystallographic c-axis. (b) The two-dimensional faceted supramolecular structure along the ac-plane.

Export to PPT

3.1.3

3.1.3 Structural description of [Ln(2,4-DFBA)₃(phen)]₂, (Ln = Tb(3), Dy(4), Ho(5))

Fig. 4a shows the structural unit of complex 4, which can be seen as a binuclear molecule consisting of six 2,4-difluorobenzoate ligands and two o-phenanthroline ligands in coordination. The coordination environments of the two central Dy(III) are the same, and the coordination number is 8. The coordination environment is obtained as a biaugmented trigonal prism column after the calculation (Madanhire et al., 2020) of the shape software (Fig. 4b). Each central Dy(III) forms a coordination environment with two N atoms in one o-phenanthroline ligand (N1, N2) and six O atoms in five 2,4-difluorobenzoate ligands (O1, O2, O3, O4, O5, O6). A closer look reveals that the coordination of oxygen atoms is divided into two types, which are double-dentate chelation (O5, O6) and bridging double-dentate (O1, O2, O3, O4). The two central Dy(III) are connected by four 2,4-difluorobenzoate ligands (O1, O2, O3, O4, O1#, O2#, O3#, O4#) in a bridging double-dentate coordination mode. The bond lengths of Dy1-O bonds range from 2.273(5)-2.430(5) Å with an average bond length of 2.350(5) Å. The average bond length of Dy1-N is 2.546(5) Å. This is similar to the previously reported (Casanovas et al., 2021) ranges for the same species of chemical bonds. To further investigate the crystal structure, the one-dimensional chain-like structure of complex 4 along the crystallographic c-axis direction (Fig. 5a) and the two-dimensional faceted (Fig. 5b) supramolecular design along the crystallographic bc-plane direction was drawn. As shown in the figure, the adjacent structural units form a one-dimensional chain in the crystallographic c-axis direction by C-H···F hydrogen bonding. The hydrogen bonding force is formed by the F4 in the bridging bidentate coordination of 2,4-difluorobenzoate and the C(20)H(20) in the bidentate chelate coordination in the adjacent structural unit with a weak force distance of 3.143 Å. The adjacent one-dimensional chains are in turn interconnected by C-H···F hydrogen bonding, forming a two-dimensional faceted supramolecular structure on the bc surface. Both hydrogen bonding interactions are created by the forces of C(20)H(20) in the 2,4-difluorobenzoate ligand using the bidentate chelate coordination mode and F(4) in the 2,4-difluorobenzoate using the bridging bidentate coordination mode. The distances were 3.143 and 3.406 Å, respectively. In analogy to the previously reported (Du et al., 2021) [Ho(2-FBA)₃(5,5‘-DM-2,2‘-bipy)]₂, the hydrogen bond length is in the appropriate range.

Close

Fig. 4

(a) Structural unit of complex 3, 4 and 5 (Ln = Tb, Dy, Ho). (b)The polyhedral coordination environment of the central Ln(III) (Ln = Tb, Dy, Ho).

Export to PPT

Close

Fig. 5

(a) The one-dimensional chain structure along the crystallographic c-axis. (b) The two-dimensional faceted supramolecular structure along the bc-plane.

Export to PPT

3.2 Infrared and Raman spectra

IR and Raman spectra are shown in Fig. 6. It can be observed from the figure that the spectra of the two ligands are very different from those of the five complexes, indicating the successful synthesis of the new complexes. And the spectral morphologies of the five complexes were the same, indicating similar structures.

Close

Fig. 6

(a) Infrared spectra of complexes 1–5 and ligands. (b) Raman spectra of complexes 1–5 and ligands.

Export to PPT

Supporting information Table S6 shows the IR and Raman spectral data. The absorption peaks of the 2,4-difluorobenzoate located at ν_C=O (IR: 1691 cm⁻¹, R: 1634 cm⁻¹) completely disappeared after the formation of the complexes. In contrast, the spectrograms of the complexes showed new absorption peaks ν_as(coo^–₎ (IR: 1616–1610 cm⁻¹, R: 1620–1614 cm⁻¹) and ν_s(coo^–₎ (IR: 1431–1427 cm⁻¹, R: 1416–1415 cm⁻¹), which indicates the involvement of oxygen atoms in the formation of coordination sites in the acidic ligands. The differences between the bands (Δ = ν_as(coo^–₎ - ν_s(coo^–₎) reflect the coordination pattern (Deacon and Phillips, 1980) of the carboxylate with Ln(III). The spectral band differences of all the complexes were in the range of 189–179 cm⁻¹, indicating that the coordination modes were chelate and bridging mode. Furthermore, the appearance of the characteristic absorption peak of ν_Ln-O in the complex spectrogram (IR: 419–417 cm⁻¹, R: 423–417 cm⁻¹) confirms the above view (Wang et al., 2015). The ν_C=N absorption peak of o-phenanthroline at (IR: 1560 cm⁻¹, R: 1507 cm⁻¹) shifts the wavenumber to the high wavenumber region after the formation of the complex, and a blue shift occurs, representing lower energy required for the vibration and a more stable complex moiety is formed (Hermansson et al., 2009). The characteristic absorption peak of ν_Ln-N (R: 242–240 cm⁻¹) was also observed in the Raman spectrum. This further suggests that the lanthanide atoms form coordination bonds with the nitrogen atoms in the o-phenanthroline (Ye et al., 2010).

3.3 Powder X-ray diffraction

Experiments with X-ray powder diffraction at room temperature were conducted on complexes 1–5 and their ligands. The two different structures of complexes 1 and 4 are used as an introduction. It can be seen in Fig. 7 (a) that the peak patterns of the complex powders 1–5 are the same, indicating that these five complexes have the same molecular structure general formula. The single crystal CIF data of complexes 1 and 4 were fitted to obtain the corresponding fitted plots Simulated 1 and Simulated 4. As shown in Fig. 7 (b), the peak pattern of the complex is different from that of the two ligands, indicating that the complex forms a new phase rather than a simple summation of the two ligands (Wu et al., 2019). In addition, the diffraction peaks of each complex powder are consistent with the data of their single-crystal CIF fits, which further indicates that the obtained complex powders have the same structure as their corresponding single crystals and the complexes are relatively pure (Zhu et al., 2019).

Close

Fig. 7

(a) XRD curves of complexes 1–5. (b) XRD curves of ligands and complexes 1 and 4.

Export to PPT

3.4 Thermal analysis

Thermal stability is a measure of the physicochemical properties of substances. Supporting information Fig. S1 shows the curves of the thermal decomposition process of the five complexes. The collected thermal decomposition data are shown in supporting information Table S7. The thermal decomposition processes of complexes 1–3 are consistent, and complex 1 is used as an example to introduce them in detail. From Fig. S1 (a), it can be seen that the complex is decomposed in two steps. The first decomposition step occurred in the temperature range 494.15–691.15 K with a weight loss of 60.58%, corresponding to the loss of two o-phenanthroline ligands and part 2,4-difluorobenzoate ligands. The second decomposition step occurred within 691.15–1009.15 K with a weight loss percentage of 18.21%, corresponding to the loss of the remaining 2,4-difluorobenzoate. The final decomposition product was La₂O₃, and the experimental total weight loss for the two-step decomposition was 78.79%, which was similar to the theoretical total weight loss of 79.39%. The thermal decomposition process of complexes 4 and 5 are similar and accomplished in three steps. 2,4-difluorobenzoate is lost in multiple stages, which is different from complexes 1–3. The final decomposition products are the corresponding metal oxides (Zhu et al., 2019).

During the thermal decomposition, the infrared escape gases of the five complexes accumulate in three-dimensional space (Fig. S2). The strongest signal peaks in the 3D stacking diagram at each decomposition stage were solved. The IR spectra of the strongest signal peaks at different temperatures were plotted and displayed in supporting information Fig. S3.

Complexes 1–3 have two stacking peaks, and complexes 4 and 5 have three stacking peaks, which is consistent with the thermal decomposition process. Take complexes 1 and 4 as examples. For complex 1, the strongest signal of the first decomposition phase was observed at 619.15 K. The characteristic absorption peaks of small organic fragments of o-phenanthroline ligand decomposition were observed, which were ν_C=N at 1611 cm⁻¹, ν_C-N at 1232 and 1257 cm⁻¹, ν_C-H at 3188–3083 cm⁻¹, and γ_C-H (772, 1050, 1130 cm⁻¹). In addition to this, a CO₂ bending vibration peak at 672 cm⁻¹ and an asymmetric stretching vibration peak at 2366 cm⁻¹ was also observed. This is due to the partial decomposition of 2,4-difluorobenzoate. The strongest peak of the signal in the second decomposition stage at 732.15 K. The bending vibration peak, and the asymmetric stretching vibration peak of carbon dioxide (672, 2361 cm⁻¹) are visible in the figure, while the vibration peak of water molecules is not obvious, which may be due to the small amount of production that the instrument cannot detect. The strongest signal peak of complex 4 (Fig. S3d) appears at three temperature nodes, 621.15 K, 773.15 K, and 820.15 K, respectively. The absorption peak of small organic fragments of o-phenanthroline decomposition was observed in the first stage. For example, ν_C=N (1607 cm⁻¹), ν_C=C (1438, 1489 cm⁻¹), ν_C-N (1231, 1257 cm⁻¹), ν_C-H (3453–3032 cm⁻¹) and γ_C-H (768, 1046, 1126 cm⁻¹). At the same time, characteristic absorption peaks of the fugitive gases H₂O (3883–3580 cm⁻¹) and CO₂ (671, 2357–2311 cm⁻¹) were also observed, which are due to partial decomposition of 2,4-difluorobenzoate ligands. The strongest signal peaks of the second and third steps are the same, with temperatures of 773.15 K and 820.15 K, H₂O (3753–3664 cm⁻¹) and CO₂ (671, 2361–2311 cm⁻¹) were observed, proving the loss was of the remaining 2,4-difluorobenzoate. In summary, the products of each decomposition stage corresponded to the results obtained from the thermogravimetric analysis.

3.5 Luminescence spectrum

Fig. 8 illustrates that the emission spectrum of Tb³⁺ at 353 nm as the excitation wavelength shows four distinct characteristic emission peaks (Zhao et al., 2004). They are ⁵D₄→⁷F_J (J = 5, 4, 3, 2), respectively located at 490, 544, 586, 623 nm. Due to the appearance of the polycrystalline state of complex 3, the presence of the crystal field leads to the splitting of the characteristic ⁵D₄→⁷F₅ emission peak located at 544 nm. Meanwhile, the ⁵D₄→⁷F₅ characteristic emission peak is the most substantial jump peak, which is why the green fluorescence (Xiao et al., 2017) of complex 3 under UV light. The green area in the CIE coordinates (0.3527,0.5793; Fig. 8, inset) also supports this conclusion. Due to the strong fluorescence performance of Tb³⁺, the double exponential function equation (Lee et al., 2010) I(t) = B₁exp(-t/τ₁) + B₂exp(-t/τ₂) was fitted, and the fitted equation τ=(B₁τ₁² + B₂τ₂²)/(B₁τ₁ + B₂τ₂) was used to calculate the fluorescence lifetime of complex 3 as 1.693 ms (Fig. 10). For complex 4, a distinct broadband absorption peak appears in the excitation spectrum, which is due to the π-π* electron leap of the organic ligand. Two characteristic leap peaks of Dy³⁺ can be observed (Carter et al., 2014) in the emission spectra. They are the characteristic transition peak of ⁴F_9/2→⁶H_15/2 located at 481 nm and the characteristic leap peak of ⁴F_9/2→⁶H_13/2 located at 574 nm, respectively. Complex 4 is located in the yellow area at coordinates (0.3775,0.4269; Fig. 9, inset).

Close

Fig. 8

Fluorescence spectrum of complex 3.

Export to PPT

Close

Fig. 10

Fluorescence lifetime decay curve of complex 3.

Export to PPT

Close

Fig. 9

Fluorescence spectrum of complex 4.

Export to PPT

3.6 Energy transfer of ligands and lanthanide ion

The phenomenon of luminescence of lanthanide complexes can be explained by the energy transfer between ligands and lanthanide ions. Energy transfer between the ligand and the lanthanide ion depends on the degree of matching between the triplet excited state and the resonance energy level of the lanthanide ion. According to Reinhoudt's rule of thumb, if the difference between the lowest singlet and lowest triplet energy levels of the same ligand is more excellent than 5000 cm⁻¹, it can be considered an effective inter-system scramble process (Segall et al., 2002). According to Latvia's rule (Latva et al., 1997) of thumb, the lowest triplet state of the ligand should be 5000–2000 cm⁻¹ higher than the lowest emission energy level of the lanthanide ion for optimal energy transfer.

The energy level calculations of the ligands are performed on a computer using the Gaussian09 series program package for simulations. The highest occupied molecular orbitals and the lowest unoccupied molecular orbitals of the ligands were investigated at the level of the 6-311G(d,p) group using the B3LYP method in density generalization theory (Karabacak et al., 2011), and the results are shown in supporting information Table S8. On this basis, the ligands' single and triplet state energy levels were calculated, and the obtained data are shown in supporting information Table S9. The energy levels of the o-phenanthroline were estimated to be 32,206 and 23,095 cm⁻¹, with an energy gap of 9111 cm⁻¹. Similarly, the energy levels of the acidic ligand 2,4-difluorobenzoic acid were 40,123 cm⁻¹ for the single heavy state and 28,766 cm⁻¹ for the triplet state, with an energy gap difference of 11,537 cm⁻¹. This shows that both o-phenanthroline and 2,4-difluorobenzoic acid ligands can undergo an effective interlinear scramble process. The lowest emission energy levels (Dejneka Matthew et al., 2003, Tsaryuk et al., 2017) of Tb(III) and Dy(III) are usually considered to be 20,430 cm⁻¹ and 21,300 cm⁻¹, and the energy differences between the lowest triplet state of the ligand and the lowest emission energy level of the Ln(III) ion are 2665 and 1795 cm⁻¹, respectively. This suggests that o-phenanthroline is an effective molecular fluorescence sensitizer that enhances the fluorescence intensity of Tb(III).

3.7 Molar heat capacity properties and thermodynamic function calculations

As a result of differential scanning calorimetry (DSC) and specific heat comparison, the molar heat capacities of complexes 4 and 5 were determined in the temperature range of 260.15 to 386.15 K. As shown in Fig. 11, the molecular vibrational energies are similar because the structures of the two complexes are similar, and their molar heat capacity curves almost overlap. With the increase in temperature, the curves increased smoothly without interruptions. No phase change peaks appeared, indicating that the two complexes were structurally stable in the temperature range of 260.15–386.15 K without any phase change, conjugation, or decomposition processes. Again, this indicates that the substance is more stable in this temperature interval. In the experimental temperature range, the temperature was fitted using the equation x = [T-(T_max + T_min)/2]/[(T_max-T_min)/2] to obtain the folded temperature x. The experimentally obtained values of Cp,m were subsequently used as y, a polynomial fit to the folded temperature x yields polynomial equations for the two complexes.

Close

Fig. 11

Curves of the average molar heat capacity of complexes 4 and 5 with temperature.

Export to PPT

Complex 4:[Dy(2,4-DFBA)₃(phen)]₂.

C_p,m/J·mol⁻¹·K⁻¹ = 1403.29077 + 231.50551x-8.36579x² + 42.15143x³-1.16524x⁴-28.57373x⁵ + 3.36235x⁶.

R² = 0.99993 SD = 1.29285.

Complex 5:[Ho(2,4-DFBA)₃(phen)]₂.

C_p,m/J·mol⁻¹·K⁻¹ = 1406.64493 + 235.43261x-9.76328x² + 26.76058x³ + 0.98861x⁴-14.59546x⁵ + 3.08112x⁶.

R² = 0.9995 SD = 1.11518.

R² is the correlation coefficient, and SD is the standard deviation.

The polynomial equation obtained above is brought into the following thermodynamic equation. $$H_T-H_{298.15K}=\underset{298.15K}{\overset{T}{\int }}{C}_{p.m}dT$$ $$S_T-S_{298.15K}=\underset{298.15}{\overset{T}{\int }}{C}_{p,m}{T}^{-1}dT$$

The reference temperature was set at 298.25 K. The values of smooth molar thermal melting, and thermodynamic functions of the two complexes were calculated at 3 K intervals. The data are shown in supporting information Table S10 and Table S11.