Introduction

Lanthanide complexes have become a research hotspot for their unique structure and excellent luminescent intensity (Kong et al. 2013; Sun et al. 2020). In addition, more and more attention has been paid to its applications in magnetic materials (Dobies Návia et al. 2018), luminescent material (Shi et al. 2021), catalyze (Qi et al. 2019), cell imaging material (Kobyłka et al. 2020), biology (Bai et al. 2021). According to the theory of soft and hard acid–base, lanthanide ions, as hard acids, have strong affinity with hard donors containing oxygen and nitrogen atoms (Su et al. 2012). The results show that the complexes with two ligands not only have more stable structures (Kariaka et al. 2016) but also enhance their fluorescence and biological properties (Zapała et al. 2018). Therefore, oxygen-containing aromatic carboxylic acids and nitrogen-containing heterocyclic ligands were the first choice for lanthanide complexes (Ji et al. 2021). The –COOH group in the aromatic carboxylic acid ligand has many binding sites, which not only meets the requirement of high coordination number of rare-earth ions but also can be coordinated by a variety of coordination modes. It was conducive to the construction of novel complexes (Zhang et al. 2021; Xiong et al. 2021). In addition, oxygen atoms in carboxylic acid ligands were easy to form hydrogen bonds with adjacent structural units, which makes the structure of the complexes more stable. The addition of nitrogen-containing heterocyclic ligands can not only meet the requirements of high-coordination number, but also improve the rigidity and stability of the complexes (Zong et al. 2015). Lanthanide ions have weak absorption of light due to the prohibition of f-f electron transition. Organic nitrogen-containing ligands can be used as "antennas" to transfer the absorbed light energy to lanthanide ions. It can promote the transfer of electrons between orbits (Ahmed et al. 2010; Xu et al. 2020) and make it has good luminous performance. This was the theory of "antenna effect" (Guo et al. 2021; Pham et al. 2021). Meanwhile, long fluorescence lifetime, narrow emission peak, and high intensity were also the advantages of lanthanide organic complexes (Zhao et al. 2021). The thermal stability and decomposition mechanism of complexes have become the point of chemists and physicists (Nascimento et al. 2019). TG-DSC/FTIR can be applied to detect the thermal behavior of the complexes at different temperatures and provide useful information for the structures of the complexes (Chesnokov et al. 2018; Zhou et al. 2021). Meanwhile, the thermal decomposition mechanism of complexes also provides an important reference for the synthesis of functional materials with certain thermal stability (Ren et al. 2020; Wang and Zhang 2020; Wang et al. 2016). In recent years, people have carried out in-depth research on organic ligands containing nitrogen and oxygen atoms and having many functions. For example, some lanthanide complexes constructed by aromatic carboxylic acids and nitrogen-containing ligands have been reported in our research group, and a series of characterization and properties of these complexes have been studied. The selected aromatic carboxylic acids were mainly benzoic acid, which was mainly because the oxygen atoms on the carboxylic group have strong affinity and various coordination modes with lanthanide ions. Nitrogen-containing heterocyclic ligands were mainly o-phenanthroline and pyridine, such as 2,2 ': 6',2 '-tripyridine, 5,5' -dimethyl-2,2 '-bipyridine and 4, 4-bipyridine, etc. (Du et al. 2021; Zhao et al. 2018). Their large conjugated structures can not only enhance the stability of complexes but also form multi-dimensional supramolecular structures. Therefore, the synthesized complexes have potential applications in sensors, safety devices, and biological imaging (Di et al. 2019; Yang et al. 2017). In this paper, we selected 2, 6-dimethylbenzoic acid and 1, 10-phenanthroline as ligands to synthesize three lanthanide complexes, namely: [Eu(2,6-DMBA)3(phen)]2(1), [Gd(2,6-DMBA)3(phen)]2·1.5H2O(2) and [Tb(2,6-DMBA)3(phen)]2·2.5H2O(3) were synthesized with 2,6-DMBA as acidic ligand, phen as neutral ligand and Ln (NO3)3·6(H2O) as a central ion. The complexes were characterized by elemental analysis, R and IR spectroscopy, and single-crystal X-ray diffraction. The thermal decomposition process and escaping gas of the complexes were studied by TG-DSC and FTIR. The fluorescence properties of solid-state complexes were studied.

Experimental

Materials and reagents

The reagents and drugs used in this study were of high purity and can be used without further purification.

Instruments and experimental conditions

The content of C, H, and N were determined by a Vario EL cube element analyzer. The single-crystal data of the complexes were collected by smart-1000 (Bruker AXS) single-crystal diffractometer at room temperature. The single-crystal data of the complexes were collected on a Smart-1000 single-crystal diffractometer, the emission source was Mo-Kα (λ = 0.71073 Å) monochromatized by graphite. At the same time, the obtained data were improved and refined by SHELXS-97 program, in which the direct method was used for the resolution of non-hydrogen atom, and the total matrix least square method was used for the refinement of its coordinates and anisotropy. The disordered aqueous solvents contained in complexes 2 and 3 were solved using the SQUEEZE program in Platon software (Sheldrick et al. 2015; Boutar et al. 2019). The full matrix least square method was used for refinement. The Raman spectra of complexes and ligands were recorded with BRUKER VERTEX-70FTIR-RAMAN II and an excitation power of 300 mW. The infrared spectra of complexes and ligands in the range of 4000–400 cm−1 were determined by KBr tablet pressing method on BRUKERTENSOR27 Fourier transform infrared spectrometer. The TG, DTG, DSC, and FTIR analysis of the complexes were measured by NETZSCH STA 449 F3 synchronous thermal analyzer and BRUKER TENSOR27 Fourier transform infrared spectrometer in the simulated atmospheric atmosphere. The fluorescence spectra and lifetime of the complexes were measured by the FS5 fluorescence spectrometer.

Synthesis

0.6 mmol(0.0901 g) of 2,6-DMBA acidic ligand and 0.2 mmol(0.0376 g) of phen neutral ligand were dissolved in 7 mL of 95% ethanol solution, stir and adjust the pH value to 5.4–6.7 with 1 mol/L NaOH solution. Then dissolve 0.2 mmol(0.0892 g, 0.0902 g, 0.08700 g) Ln (NO3)3·6(H2O) in 3 mL distilled water, drop the ligand solution into the stirring lanthanide nitrate solution, and stir for 6–8 h. Stand for 12 h, filter, put the filtered liquid into the beaker, and stand for one week to obtain the crystal. Scheme 1 was the synthesis scheme of complex 1.

Scheme 1
scheme 1

Synthesis scheme of complex 1

Full size image

Results and discussion

Composition of complexes

Table 1 lists the content of element (C, H, and N) in the complexes, and the experimental value on the verge of the theoretical value, indicating that the complexes obtained in the experiment were relatively pure.

Table 1 Elemental analysis data of complexes 1–3
Full size table

Raman and infrared spectroscopy

To determine the complexes structure more accurately, Raman and IR spectra of ligands and complexes were determined. Figure 1 shows the R and IR spectra of acid ligand, neutral ligand, and complexes. The spectral data were shown in Table 2. As you can see from the picture that the characteristic absorption peak of carboxylic group νC=O in acidic ligand appears at 1639 cm−1(R), 1690 cm−1(IR), however, this peak does not appear in the complexes. The characteristic absorption peaks of νas(COO−) and νs(COO−) appear in the complexes, they belong to carboxylic acid anions, located at 1587–1602 cm−1(R), 1615–1618 cm−1(IR) and 1444–1451 cm−1(R), 1451–1457 cm−1(IR) respectively. By comparing the neutral ligand with the complexes, the characteristic peaks of phen νC=N (1564 cm−1) (R), (1602 cm−1) (IR) have an obvious shift. Indicating that the N atoms of phen ligand were coordinated with lanthanide complexes (Yao et al. 2006). The characteristic absorption peaks of Ln–O and Ln-N were discovered at 405–407 cm−1(IR), 416–418 cm−1(R), and 272–279 cm−1(IR), respectively. It shows that the oxygen atoms in 2,6-DMBA and the nitrogen atoms in phen coordinate with the central metal ion (He et al. 2014; Wang et al 2016).

Fig. 1
figure 1

R a and IR b Spectra of ligands and complexes

Full size image
Table 2 Characteristic bands(cm−1) of Raman and infrared spectra of ligands and the title complexes
Full size table

Description of crystal structure

The three complexes were divided into two types by single-crystal X-ray diffraction, the first type was complex 1, and the second type were complexes 2 and 3. Complexes 2 and 3 contain disordered solvent molecules removed by the SQUEEZE procedure. These disordered solvent molecules can be calculated by the number of electrons contained in the structural units combined with elemental analysis and thermogravimetric analysis (Boutar et al. 2019). Among them, the structural units of complexes 2 and 3 contain 18.5 e− and 36.9 e−, respectively. The results show that complexes 2 and 3 contain 1.5 and 2.5 disordered water molecules, respectively. Table 3 shows the crystallographic structure data and refined crystal structure parameters of complexes 1–3. The lengths of selected hydrogen bonds were listed in Table 4. Next, we take complexes 1 and 3 as examples to introduce the crystal structure of the complexes. Table 5 lists the hydrogen bond parameters of the three complexes.

Table 3 X-ray diffraction data and refinement details of complexes 1, 2 and 3
Full size table
Table 4 Bond lengths for complexes 1–3
Full size table
Table 5 Selected Hydrogen-Bond Lengths (A˚) and Angles (°)complexes 1–3
Full size table

[Eu(2,6-DMBA)3(phen)]2(1)

As shown in Fig. 2a, complex 1 has a unique binuclear structure and belongs to the triclinic system with space group Pī. It was made up of two Eu(III) ion, six 2,6-DMBA and two phen, and the coordination number was nine. The seven oxygen atoms O5 and O6, O3 and O4, O1, O2 and O1 'in 2,6-DMBA were coordinated by chelating bidentate, bridged bidentate and bridged tridentate, respectively. The two N atoms in the phen ligand were coordinated by bidentate chelation. The coordination environment of the complex was a distorted single cap tetragonal antiprism (Fig. 2b). The coordination mode of 2, 6-dimethylbenzoic acid was shown in Fig. 3. Comparative the bond lengths of the central atom of the complex with O and N in Table 4, the bond lengths of Eu-N and Eu–O were (2.618–2.716 Å) and (2.331–2.644 Å), respectively. It can be concluded that the average bond length of Eu and O was smaller than that of Eu and N. This was also why neutral ligands were lost before acidic ones during thermal decomposition. The adjacent structural units form a 1D chain structure along the c direction through the interaction of hydrogen bonds (3.385 Å). The 1D supramolecular structure in the c-direction forms a 2D supramolecular structure in the b-axis and c-axis through hydrogen bonding between ligands (Figs. 4a, b).

Fig. 2
figure 2

a The crystal structure of 1. b The polyhedral model of Eu3+ ion center

Full size image
Fig. 3
figure 3

Coordination mode of 2, 6-DMBA−

Full size image
Fig. 4
figure 4

a The one-dimensional chains structure along the c axis. b The two-dimensional sheet is about the bc plane

Full size image

It was different from the complex [Eu(2,6-DMBA)3(5,5'-DM-2,2'-bipy)]2 (2,6-DMBA = 2,6-dimethylbenzoic acid, 5,5'-DM-2,2'-bipy = 5,5'-dimethyl-2,2'-bipyridine) reported by Li Yingying (Li et al. 2019). It has been reported that the complex was centered on a 9-coordinated Eu(III) ion and forms a twisted tetragonal inverse prism geometry with surrounding nitrogen and oxygen atoms. The difference was that the reported complexes crystallize in the P21/c-space group of the monoclinic system. The two complexes have the same central ion and acid ligand, but the difference was that there were two benzene rings in 5,5'-DM-2,2'-bipy and three benzene rings in phen. The steric hindrance of the complexes results in different spatial crystal systems.

[Tb(2,6-DMBA)3(phen)]2•2.5H2O(3)

It was centered on a 9-coordinated Tb(III) ion and coordinates with seven oxygen atoms in the 2, 6-DMBA ligand and two nitrogen atoms in the phen ligand (Fig. 5a). Oxygen and nitrogen atoms coordinate with Tb(III) ion by chelating bidentate, bridged bidentate, and bridged tridentate respectively. The coordination environment of complex 3 was a distorted mono-capped square anti-prismatic (Fig. 5b). The average bond length of Tb-N(2.611 Å) was greater than that of Tb-O(2.421 Å), this data can be obtained from Table 4. This further shows that the coordination ability of nitrogen atom with lanthanide ion was not as good as that of oxygen atom with lanthanide ion compared with complex 1, the average bond length of Eu-N(O) was slightly longer than that of Tb-O(N), which was due to the "lanthanide contraction" (Deng et al. 2010). Adjacent structural units form 1D chains along the c-axis through weak hydrogen bond interactions (3.385 Å) between oxygen atoms in the acid ligand and carbon atoms in the neutral ligand (Fig. 6).

Fig. 5
figure 5

a The crystal structure of 3. b The polyhedral model of Tb3+ ion center

Full size image
Fig. 6
figure 6

The one-dimensional chains structure along the b axis

Full size image

The complex [Tb2L6(phen)2] (L = 4-fluorophenylselenoacetic acid, phen = 1,10-phenanthroline) was reported by Zhang Fan (Zhang et al. 2021). It has many similarities with the complex 3 synthesized in this paper. For example, it has the same coordination number, coordination environment, and coordination mode. The different was that the space group of the reported complexes was P21/c. In addition, the complex 3 synthesized in this paper contains disordered solvent water molecules. This may be due to the different number of substituents in acidic ligands, resulting in different steric hindrance, resulting in different spatial crystal systems of complexes.

Thermal analysis

The thermal behavior of the complexes with temperature was studied. Table 6 lists the thermal decomposition data obtained by simulating the air atmosphere, and the heating rate was 10 K·min−1. Figure 7 was the TG-DTG-DSC curves of complexes 1–3. Complexes 2 and 3 have the same crystal structure, so their thermal decomposition process was similar. Therefore, the thermal behavior of complexes 1 and 3 were described in detail.

Table 6 Thermal analysis data of complexes 1–3 in simulated air atmosphere
Full size table
Fig. 7
figure 7

TG-DTG-DSC curves of the complex 1 (a); complex 2 (b); complex 3 (c)

Full size image

Combined with TG and DTG analysis (Fig. 7a), complex 1 was decomposed in two steps. At 502.15 K, the decomposition of the first step begins, and the weight loss rate was 23.17%, which was close to the calculated value of losing two phen (calcd: 23.11%). This suggests that the breakdown of the first step can be attributed to the loss of phen ligands. The weight loss of the second step started at 626.15 K and ended at 974.15 K, with a weight loss rate of 53.60% (calcd: 54.32%), indicating that the decomposition of the second step could be attributed to the decomposition of six acid ligands. The total weight loss rate was 76.77%, which was in good agreement with the theoretical value(77.43%). The final product was Eu2O3. In conclusion, the Eu-N bond was destroyed prior to the Eu–O bond. The thermal decomposition process of the complex 1 was as follows:

$$ \left[ {{\text{Eu}}_{{2}} \left( {{2},{6} - {\text{DMBA}}} \right)_{{6}} \left( {{\text{phen}}} \right)_{{2}} } \right] \to \left[ {{\text{Eu}}_{{2}} \left( {{2},{6} - {\text{DMBA}}} \right)_{{6}} } \right] \to {\text{Eu}}_{{2}} {\text{O}}_{{3}} $$

DTG curve showed that complex 3 was decomposed in three steps (Fig. 7c). In the temperature range of 395.15 ~ 502.15 K, the first step weight loss rate was 2.70% (calcd: 2.78%). This step of decomposition can be attributed to the loss of 2.5 water molecules. Unlike complex 3, complex 2 loses 1.5 disordered water molecules in the first step. The second step of weightlessness occurs at 502.15 ~ 608.15 K, and the weightlessness rate was 22.82% (calcd: 22.27%). Thus, the second step of decomposition can be attributed to the loss of two neutral ligands. The last step of decomposition occurs between 608.15–938.15 K, and the weight loss rate was right around the theoretical value of 51.85% (found: 52.58%) of the loss of six acidic ligands. Combined with the above analysis, the final remaining product was 21.9% (calcd 23.1%), and the final product was Tb4O7 (Łyszczek et al. 2016; Wang et al. 2011). The thermal decomposition process of the complex 3 was as follows:

$$ \left[ {{\text{Tb}}_{{2}} \left( {{2},{6} - {\text{DMBA}}} \right)_{{6}} \left( {{\text{phen}}} \right)_{{2}} } \right]\cdot{2}.{\text{5H}}_{{2}} {\text{O}} \to \left[ {{\text{Tb}}_{{2}} \left( {{2},{6} - {\text{DMBA}}} \right)_{{6}} \left( {{\text{phen}}} \right)_{{2}} } \right] \to \left[ {{\text{Tb}}_{{2}} \left( {{2},{6} - {\text{DMBA}}} \right)_{{6}} } \right] \to {\text{Tb}}_{{4}} {\text{O}}_{{7}} $$

Evolved gas study during thermal decomposition

In order to further determine the thermal decomposition products, the three-dimensional infrared stacking diagram of the complexes in the thermal decomposition process was determined.

Figures 8a–c shows the three-dimensional infrared stacking diagram of the gas escaping from complexes 1–3 during thermal decomposition. No signal peak at the first stage was observed from the three-dimensional infrared stacking diagram of complexes 2–3, which may be due to less gas generation and a longer instrument pipeline. Therefore, complexes 1–3 have similar three-dimensional infrared stacking patterns. We take complex 3 as an example.

Fig. 8
figure 8

The 3D infrared spectrum stacking diagram of gas escaping from complex 1 (a); complex 2 (b); complex 3 (c)

Full size image

The two-dimensional infrared spectra at 489.32 K and 753.02 K were obtained from the three-dimensional infrared stacking diagram (Fig. 9c). The characteristic absorption bands of CO2 (2318–2367, 683 cm−1) and H2O (3563–3847 cm−1) were observed in the two-dimensional spectrum at a temperature of 489.32 K. Meanwhile, the absorption bands of organic molecular fragments were also detected: νC-H (3083–3401 cm−1), νC=C (1569 cm−1), νC=N (1589 cm−1), νC-N (1191 cm−1) and γC-H (764, 865, 1062 cm−1). It indicates that the phen ligand was decomposed in this process. The characteristic absorption bands of CO2 (2163–2291 cm−1, 668 cm−1) and H2O (3566–3776 cm−1) can be observed from the two-dimensional spectra at 753.02 K. It indicates that the remaining acidic ligands escape as gaseous CO2 and H2O in this step. This was in compliance with the thermal decomposition process of the above complexes.

Fig. 9
figure 9

The 2D infrared spectra of complex 1 (a); complex 2 (b); complex 3 (c)

Full size image

Luminescent property

To further probe the luminescent property of lanthanide complexes, the solid-state fluorescence spectra of complexes 1 and 3 were determined by fluorescence spectrometer. Complexes 1 and 3 emit intense red and green light respectively under UV lamp irradiation (Figs. 10, 11). The excitation spectra of complex 1 was shown in Fig. 10a. The π → π* electron transition of organic ligand makes complex 1 has a wide absorption band in the range of 250–450 nm, this further suggests that organic ligands can act as antennas to efficiently transfer absorbed energy to Eu(III) ions. However, due to the f-f configuration transition of Eu(III) ions, other weak excitation peaks appear in this wavelength range. The emission spectrum (Fig. 10b) was measured within the limits of 500–750 nm with 335 nm as the excitation wavelength. The intensity of 5D0 → 7F2 (620 nm) transition was higher than that of 5D0 → 7F0 (579 nm), 5D0 → 7F1 (590 nm), 5D0 → 7F3 (650 nm) and 5D0 → 7F4 (695 nm). This was also the reason why Eu complexes emit red light (Ridenour et al. 2017).

Fig. 10
figure 10

The solid-state excitation (a) and emission (b) spectra for 1

Full size image
Fig. 11
figure 11

The solid-state excitation (a) and emission (b) spectra for 3

Full size image

In the range of 250–450 nm, an obvious broadband absorption was observed in the excitation spectrum of complex 3 (Fig. 11a), this was mainly due to the π → π* electron transition of organic (Rybak et al. 2012). The emission spectrum of Fig. 11b was measured at 345 nm, and four characteristic transitions of Tb(III) ion can be observed within the limits of 500–750 nm. The transition intensity of 5D4 → 7F5 at 545 nm was larger than that of 5D4 → 7F6 (490 nm), 5D4 → 7F4 (590 nm) and 5D4 → 7F3 (620 nm). It was also the reason why complex 3 emits strong green light under ultraviolet light (Łyszczek et al. 2015). It has the same characteristic transition and luminescence color as the Tb complex published by Ye et al. (2010).

The emission spectra of complexes 1 and 3 were brought into the color coordinate diagram, respectively. The calculated color coordinates were (0.6633, 0.3354), (0.3425, 0.5968). Complexes 1 and 3 were located in the red and green regions, respectively (Fig. 12).

Fig. 12
figure 12

CIE chromaticity diagram of 1 and 3 corresponding to fluorescence emission spectrum

Full size image

In addition, the fluorescence lifetime of lanthanide complexes were also of great meaning in the study of physical, chemical, and biological excitation kinetics. The fluorescence lifetime decay curves of solid complexes 1 and 3 at the optimal excitation and emission wavelengths were measured (Figs. 13, 14). The fluorescence lifetime data of the complexes were fitted by double exponential function equation: I(t) = B1·exp(-t/τ1) + B2·exp(-t/τ2), and the corresponding fitting parameters B1, B2, and decay time were obtained τ1, τ2. At the same time, the fluorescence lifetimes of complexes 1 and 3 were calculated by formula τ = (B1τ12 + B2τ22)/ (B1τ1 + B2τ2) as 1.034 ms and 0.557 ms, respectively.

Fig. 13
figure 13

Fluorescence attenuation and fitting curve of 1

Full size image
Fig. 14
figure 14

Fluorescence attenuation and fitting curve of 3

Full size image

In summary, complexes 1 and 3 can emit characteristic transitions corresponding to lanthanide ions and have strong fluorescence intensity and long lifetime. Therefore, these two complexes have potential application prospects in the field of luminescent materials.

Conclusions

In this paper, three lanthanide complexes with novel structures were prepared and characterized by conventional solution method using 2, 6-DMBA, and phen as the ligands. The results show that the complexes 1–3 were all binuclear complexes with 9 coordination and all crystallize in the Pī-space group of trisoclinic crystal system. The different supramolecular structures of complexes 1 and 2–3 may be caused by different hydrogen bond interactions and lanthanide contraction effects. However, the binuclear units of complexes 1–3 form 1D chain structure through hydrogen bond interaction. TG-DSC/FTIR study showed that the three complexes had good thermal stability. Solid-state fluorescence test shows that terbium complex and europium complex emit strong characteristic green and red light, respectively. It lays a foundation for the application of lanthanide complexes in the field of luminescent materials.