1 Introduction

Nitroaromatic compounds (NACs) are important agents for environment pollution [1,2,3], which are formed by incomplete oxidation of organic matter and coexist in various industrial and agricultural discharges [3, 4]. The strong biological toxicity, high stability and poor biodegradability have greatly damaged the environment [5]. Nitrobenzene is the simplest and basic ingredient in the explosive family, and its killing effect cannot be ignored. Therefore, the development of a fast and effective method for detecting nitroaromatic compounds is a hot research topic [6, 7]. Numerous d10 metal ions, particularly Zn2+ and Cd2+ ions, have strong complexation affinities for N/O- donor atoms as well as favourable photoactive characteristics [8, 9]. Because of their fascinating framework architectures, metal–organic frameworks (MOFs) have been developed quickly in research on gas storage/separation, optics, dye capture and catalysis over the last few decades [10,11,12,13,14,15,16]. And the literature shows that MOF composed of and electron-rich ligands is more sensitive to nitro aromatic explosives [17,18,19,20]. The pore structure of MOF can be effectively adjusted by introducing various functions, such as -NH2, -CONH2, binaphthol and pyridyl, to improve the detection performance of analytes.

Based on these considerations, the tripodal 1,3,5-tris(carboxymethoxy)benzene acid (H3L) (Scheme 1) and 2,2’-bipyridine are selected as ligands. According to pH, H3L can be fully or incompletely deprotonated making possible the creation of coordination polymers with a variety of topologies. In addition, the flexibility of -OCH2- spacers in the carboxylic acid ligand facilitates diverse coordination modes with metal centers. The constructed [Cd3(L)2(bpy)2]n (1) displayed two-dimensional layer structure and further linked into three-dimensional (3D) supramolecular framework through π⋯π stacking interactions. As expected, complex 1 effectively detects electron-deficient nitro aromatic explosives and is highly sensitive to p-nitroaniline (4-NA). The result of theoretical calculation revealed that the efficient electron transfer led to high quenching efficiency of 4-NA. Thus, complex 1 effectively detects electron-deficient nitro aromatic explosives and is highly sensitive towards p-nitroaniline (4-NA), and the result of theoretical calculations revealed the efficient electron transfer mechanism and quenching efficiency of 4-NA.

Scheme 1
scheme 1

The diagram of 1, 3, 5-tris(carboxymethoxy) benzene ligand

2 Experimental

2.1 Materials and Instrumentation

The H3L ligand and other chemicals were bought through commercial means without further purification. The IR spectra of KBr pellets were captured on a Nicolet Avatar-360 spectrometer. The elemental analyses were carried out on a Flash 2000 elemental analyzer. TGA was tested in a nitrogen atmosphere using the SDT 2960 thermal analyzer. By using the UV-2600 UV/VIS spectrophotometer and the CARY Eclipse fluorescence spectrophotometer, solid UV and fluorescence measurements were taken, respectively. Density functional theory (DFT) is used to do the theoretical computation with a B3LYP/6–31 + G* accuracy level using the Gaussian 09 W software package.

2.2 Synthesis of [Cd3(L)2(bpy)2]n(1)

H3L (0.025 mmol, 7.5 mg), Cd(CH3COO)2·2H2O (0.05 mmol, 13.33 mg), and 2,2’-bipy (0.05 mmol, 7.80 mg) were added into N,N-dimethylacetamide (3 mL) and H2O (2 mL). The mixture was placed in a Teflon-lined autoclave, heated at 120 oC for 72 h, before cooling to room temperature. Colorless block crystals of 1 were obtained. Anal. Calcd for C44H34Cd3N4O18: C 42.48, H 2.75, and N 4.50; Found: C 42.51, H 2.72, and N 4.53. IR/cm− 1 (KBr): 1597(s), 1421(m), 1323(m), 1152(s), 1063(m), 836(w), 764(s), 716(m), 647(s), and 587(s) of (s).

2.3 Photoluminescence Experiment

In the fluorescence quenching experiment, powder sample (3 mg) of 1 was ultrasonically dispersed in deionized water (3 mL) for 30 min. At the excitation of 285 nm, the newly prepared nitro explosive solution (100 ppm) was added incrementally, and then measured the fluorescence.

2.4 X-ray Crystallography

Diffraction data of complex 1 were collected on an Oxford Diffraction SuperNova diffractometer outfitted with graphite-monochromated Mo-Kα radiation (λ = 0.71073 Å) [21]. CrysAlisPro was used to data collection and processing. SADABS program was utilized for an empirical absorption correction [22]. The structure was solved by direct methods and refined using full matrix least squares on F2 using the SHELXTL software [23]. Table 1 displays crystallographic data of complex 1 and Table S1 showed a selection of bond lengths and angles.

Table 1 Crystallographic data for complex 1

3 Results and Discussion

3.1 Description of Crystal Structure of 1

The asymmetric unit of complex 1 contains two Cd(II) atoms, a completely deprotonated acid and a 2,2’-bipy (Fig. 1a). Both metal ions show an octahedral geometry. Cd1 is chelated by 2,2’-bipy ligand and four O atoms from three symmetry related L3− ligands in a highly distorted octahedron, while Cd2 located on a center of symmetry is surrounded by six O atoms from symmetry related L3− ligands. The fully deprotonated acid ligand displays two of µ211 and one of µ212 coordination modes. Cd1-N bond distances are comparable, of 2.338(4) and 2.347(4) Å, while the Cd-O bonds lengths range from 2.189(3) to 2.452(3) Å. The N/O-Cd1-N/O angles range from 53.77(10) to 154.75(13)°, while O-Cd2-O values are close to an almost regular octahedron. As illustrated in Fig. 1b and c, the crystal packing shows a two-dimensional layered structure comprising trinuclear Cd1-Cd2-Cd1’ cluster connected by carboxylic ligands, where the Cd1-Cd2 distance is 3.819(6) Å [24].

Fig. 1
figure 1

(a) Coordination environment of Cd(II) in complex 1. (b, c) Two-dimensional layers of complex 1. (d) 3D network linked by π···π stacking interactions. Symmetry codes: # 1: x, - y + 1, z – 1/2; # 2: x, y -1, z; # 3: - x, y − 1, - z + 1/2; # 4: – x, – y + 1, - z; # 5: - x, - y, - z

3.2 Characterization and Fluorescence Stability Study of 1

The experimental results of PXRD diffraction show that the main powder diffraction peak positions of complex 1 are basically consistent with that of theoretical simulation, suggesting the high phase purity of complex 1 (Fig. 2a).

Owing the strong fluorescence of d10-based coordination polymers, the solid-state fluorescence of carboxylic acid ligand and compound 1 were measured (Fig. S1). Strong emission of carboxylic acid ligand can be seen at 302 nm (λex = 260 nm), while strong emission of complex 1 can be observed at 438 nm (λex = 285 nm). A clear red-shift of 136 nm can be observed comparison to free carboxylic acid ligand, which may result from the electron transfer triggered by the binding of ligand to metal ion [25]. To investigate the fluorescence stability of complex 1, the sample was immersed in water with different pH value (pH = 3–12). From Fig. 2b, the fluorescence intensity of complex 1 only have small change at different pH value, suggested that complex 1 has good chemical and fluorescence stability.

Fig. 2
figure 2

(a) The PXRD of complex 1. (b) Fluorescence response of complex 1 at different pH

4 Photoluminescence

To explore the quenching efficiency of 1 to NACs, the fluorescence quenching experiments were carried out by increasing the amount of explosive in the water dispersion of complex 1. Five nitroaromatic compounds (Scheme S2), p-nitroaniline (4-NA), o-nitroaniline (2-NA), p-nitrophenol (4-NP), 3-nitrophenol (3-NP) and m-dinitrobenzen (1,3-DNB), were chosen to investigate the quenching efficiency of 1. For each luminescent sensing study, the grounded sample of 1 (3 mg) was added into a 3 mL aqueous solution containing nitroaromatic compound. The findings revealed that complex 1 had diverse fluorescence quenching effects on five distinct nitrobenzene compounds (Fig. 3a). The quenching efficiency of 4-NA (100 ppm) is up to 95.37%, much more than other NACs. The quenching efficiency of 2-NA and 4-NP is 87.06% and 85.64%, respectively.

Fig. 3
figure 3

(a) Fluorescence intensities of complex 1 dispersed in water solution with various NACs (100 ppm). (b-d) Relative emission peak intensity of complex 1 in the presence of different concentrations of NACs. Inset: Stern − Volmer plot of complex 1

To test the influence of complex 1 on the detection of 4-NA, 2-NA, and 4-NP in detailed, the quantitative analysis experiments were performed. The prepared NACs solution (20 µL, 100 ppm) was mixed with powder sample 1 (3 mg) ultrasonically dispersed in distilled water (3 mL). As depicted in Fig. 3b and d, the fluorescence intensity declines with incremental addition of 4-NA, 2-NA and 4-NP solution. The equation I0/I = Ksv [A] + 1 was utilized to calculate Sterne-Volmer (SV) data of complex 1 for 4-NA, 2-NA and 4-NP solution, where I0 and I are the luminous intensities before to and following analyte incorporation, respectively. [A] is the analyte concentration, and Ksv is the Sterne-Volmer constant (M− 1) [26]. A strong linear relationship between the concentration of 4-NA, 2-NA, and 4-NP and the luminous intensity may be seen in the low concentration range. The SV plot demonstrates that the Ksv values of complex 1 for 4-NA, 2-NA and 4-NP are 4.88 × 104 M− 1, 3.82 × 104 M− 1 and 4.10 × 104 M− 1, respectively. The relative limits of detection (LODs) for 4-NA, 2-NA, and 4-NP are 5.78 × 10− 6 M, 7.69 × 10− 6 M, and 7.17 × 10− 6 M, respectively, based on the 3δ/Ksv ratio (where δ is the standard deviation of three times luminous intensity in the blank solution). These results indicated that complex 1 has high sensitivity to nitroaromatic compounds (Table S2).

5 Mechanisms for the Fluorescence Quenching

In order to better understand high sensitivity of complex 1 to NACs, the fluorescence quenching mechanism were explored by theoretical calculation and the UV absorption spectra. The capacity of nitro compounds to absorb electrons allows electrons to be transported from the electron-rich skeleton to NACs. The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of ligand and NACs were verified by DFT (Fig. 4a). The LUMO energy of H3L ligand is -0.6133 eV, which is higher than that of NACs explosive. Different degrees of fluorescence quenching may arise from the excited electrons in the high LUMO of H3L being transported to the low LUMO of NACs. The calculated LUMO energy values of NACs are in the order: 4-NA < 2-NA < 4-NP < 3-NP < 1,3-DNB (Fig. 4b), which are consistent with the experimental quenching efficiency. According to the computed outcomes, the 4-NA’s excellent quenching efficiency may be attributed to an effective electron transfer.

Fig. 4
figure 4

(a) HOMO and LUMO energy levels of H3L ligand and the selected NACs calculated by density functional theory (DFT). (b) HOMO and LUMO energies of the H3L ligand and NACs. (c) Spectral overlaps between the absorption spectra of the NACs and the emission spectra of complex 1

In order to further confirm this fluorescence quenching mechanism, the UV-Vis absorption spectra of NACs were tested. The efficient overlap between the NACs’ absorption band and the MOFs’ emission band, as reported in the literature [2, 3], is advantageous for improving the resonance energy transfer’s effectiveness in quenching fluorescence. As depicted in Fig. 4c, NACs, particularly 4-NA, showed a strong overlap with complex 1’s emission band, which was consistent with the fluorescence quenching experiment and theoretical calculation. Therefore, the mechanism of fluorescence quenching may be a synergistic effect of electron transfer mechanism and resonance energy transfer mechanism between complex 1 and NACs [9, 27].

5.1 Stability Analysis

After the fluorescence quenching experiment, the samples dispersed in the solution were collected separately, and washed three times with methanol and distilled water, then placed at room temperature until the water in the sample was completely evaporated to study the PXRD. As shown in Fig. S2, complex 1 still maintained its own MOF structure before and after the fluorescence quenching measurement, indicating the good stability.

6 Conclusion

In summary, a three-dimensional supramolecular network, [Cd3(L)2(bpy)2]n (1), has been designed and synthesized. Complex 1 has not only high selectivity and sensitivity for nitroaromatic explosive, but also good chemical stability. The outcomes showed that complex 1 can be a potential fluorescent probe for nitroaromatic chemical detection. Overall, this work has important reference value and significance for practical application.