1 Introduction

With the vigorous development of chemical industry, the discharged pollutants produced by factories are increasing. Metal ions and nitro-aromatic explosives (NAEs) in the pollutants have aroused widespread concerns because of their threat to ecological environment and public security [1,2,3]. Many metal ions, especially to highly-charged metal ions, which could cause irreversible damage to living organisms and ecosystem if these metal ions are discharged into the environment, and the Fe3+ ion is also considered as toxic pollutant when the concentration is too high [4,5,6]. Nitro-aromatic explosives (NAEs) (trinitrophenol, p-nitrophenol, 4-nitrobenzene) are highly toxic chemicals, which exist in industrial production such as dyes, pharmaceuticals, explosives and paints, have irritating effects on eyes, respiratory tract and skin, and threaten people's life and health [7,8,9,10]. In addition, NAEs are strong oxidants, which react violently with reducing agents to cause explosions, which are very harmful to society [11,12,13,14]. Therefore, the effective detection of NAEs is of great significance to people’s production and life. At present, the traditional methods for detecting ions/molecules mainly include mass spectrometry, gas chromatography, high performance liquid chromatography, raman spectroscopy, ion mobility spectrometry and atomic absorption spectrometry. However, these traditional detection methods have the disadvantages of complicated operation, inconvenient equipment and high detection cost [15,16,17]. Therefore, it is very necessary to develop a fast, convenient, effective and cheap detection technology to detect metal ions and NAEs. These problems are still great challenges for scientific researchers.

In recent years, researchers have adopted some porous materials as sensors for detecting various metal ions and NAEs. Re-MOF (rare earth metal–organic framework) is a kind of porous skeleton material formed by self-assembly of rare earth metal nodes (ions, chains or clusters) and organic ligands. This kind of material not only has high porosity, adjustable pores diameter, diverse topological structure, but also has unique luminescent advantages, such as high color purity, large stokes shift, high quantum yield, narrow emission band [18, 19]. Due to the advantages of high sensitivity, fast response and real-time monitoring, Re-MOFs have become more and more attractive in the field of detection, many Re-MOFs have been developed into luminescent sensors for various harmful pollutants in the environment [20,21,22].

Up to now, researchers have reported some Re-MOFs for fluorescence detection of metal ions and NAEs. Liu et al. reported a series of isomorphic Re-MOFs, which have high sensitivity and selectivity for Fe3+, 2,4-DNP and 4-NP, respectively, and can be quickly detected and identified [23]. Wu et al. successfully constructed a multi-response fluorescence sensor (Ga-MOF), which can effectively detect Fe3+ and nitrobenzene with high selectivity and sensitivity [24]. Zhang et al. reported a series of Eu3+/Tb3+-MOFs, all of the compounds were developed as efficient fluorescence “turn-off” sensors to detect Fe3+ and 4-NP in DMF solutions [25]. The above results prompted us to continue to explore new luminescent Re-MOFs as candidate materials for multifunctional sensors.

On the other hand, the relatively soft tetracarboxylic acid ligand (1,1'-butadiynebenzene-3,3',5,5'-tetracarboxylate) (H4BBTC) can make two aromatic rings rotate freely because of its butadiyne group. This feature is helpful to construct MOFs with novel topological structure and multifunctional properties. Some MOFs based on this ligand have shown good activity in gas adsorption and magnetic applications [26,27,28]. More importantly, in view of the large conjugated system of H4BBTC ligand, it is expected that this ligand can be used to construct MOFs with good fluorescence activity. However, as far as we know, the application of MOFs based on H4BBTC ligand in the fluorescence detection of Fe3+ and NAEs has not been reported so far.

Based on the above considerations, and in continuation to our previous investigations on construction of MOFs from butadiynebenzene–carboxylate ligand, two isostructural 2D Re-MOFs [Re = Y (1) and Ho (2)], were successfully synthesized through solvothermal method. Re-BBTC were assembled with dinuclear Re2(CO2)2 SBU units and BBTC4− ligands, both structures were distinct from other known Cu(II) and Eu(III) MOFs with the same ligand [29, 30]. Notably, complex 1 exhibited high sensitivity and good selectivity towards Fe3+ and TNP, respectively. The fluorescence quenching may be attributed to the resonance energy transfer between analytes and skeletons.

2 Experimental

2.1 Materials and Measurements

1,1'-butadiynebenzene-3,3',5,5'-tetracarboxylate (H4BBTC) was prepared according to the method in the previous literature [29, 30]. YCl3⋅6H2O and HoCl3⋅6H2O were purchased from chemical company and used directly. All reagents and materials are of analytical grade and used without further purification.

The measurement of powder X-ray diffraction (PXRD) was carried out at room temperature with Cu Kα (λ = 1.54056 Å) radiation at a scanning speed of 5° min−1 and a step size of 0.02° in 2θ. The elemental analysis (C, H and N) was performed on Perkin-Elmer 240 elemental analyzer, and the IR spectrum was measured with KBr pellets on a NICOLET iS10 spectrometer. Using DTA-TGA 2960 thermogravimetric analyzer, thermogravimetric analysis (TGA) was performed at a heating rate of 20 °C min−1 under N2 atmosphere, and photoluminescence spectra were recorded at room temperature with Hitachi F4600 fluorescence spectrometer. UV–Vis spectra were collected on PerkinElmer Lambda 365 UV–Vis spectrophotometer.

2.2 Synthesis of Complexes 1 and 2

[Y2(BBTC)(DMF)5Cl2]·2DMF·H2O (1). A solution of H4BBTC (5 mg, 0.013 mmol), YCl3⋅6H2O (8 mg, 0.026 mmol), DMF (0.4 mL) and HCl (0.03 mL, 1 M in DMF) were mixed and sealed in a 15 mL Teflon–lined autoclave and heated to 85 °C for 24 h. Colourless block–shaped crystals were obtained after gradually cooled to room temperature at a rate of 5 °C h−1 (yield: 65% based on Y). Anal. calcd for C41H57Cl2Y2N7O16: C, 42.74; H, 4.98; N, 8.51. Found: C, 42.82; H, 4.76; N, 8.71. Selected IR data (KBr pellet, cm−1): 3269 (b), 3071 (w), 2933 (w), 1650 (s), 1554 (m), 1435 (s), 1379 (s), 1250 (m), 1110 (s), 779 (s), 735 (s), 716 (s).

The analogous procedure above–mentioned was used for preparation of complex 2 (see ESI†).

2.3 Crystal Structure Determination

X-ray analyses of single crystals were carried out on Bruker Smart Apex II CCD diffractometer using graphite monochromatic Mo/Kα radiation (λ = 0.71073 Å) at 296 K. Using SAINT and SADABS software packages for data reductions and absorption corrections [31]. By using the direct method of SHELXL − 97 software package [32], the structures were solved. All hydrogen atoms were placed at the calculated positions and refined on the parent atoms. All non-hydrogen atoms were anisotropically refined by full − matrix least − squares method on F2. In complexes 1 and 2, solvent molecules in the structures were highly disordered, so it was impossible to refine them by using the traditional discrete–atom models. In order to solve this problem, the contribution of solvent electron density was removed by SQUEEZE routine in PLATON [33]. The main data of collection and refinement details of complexes 1 and 2 are given in Table 1. The selected bond lengths and bond angles are listed in Tables Sl and S2†. The CCDC reference numbers are 2,308,970 for complex 1 and 2,308,971 for complex 2, respectively.

Table 1 Crystallographic data and structural refinements for 1 and 2
Full size table

3 Results and Discussion

3.1 Crystal Structures

Single crystal X–ray diffraction analyses reveal that complexes 1 and 2 are isomorphous, exhibit similar 2D frameworks. Besides this, the IR spectra of complexes 1 and 2 are almost identical (Fig. S1†). As a result, only the structure of [Y2(BBTC)(DMF)5Cl2]·2DMF·H2O (1) is representatively described in detail. The structure of complex 2 is shown in Fig. S2†.

Crystal 1 belongs to monoclinic system, and its space group is P21/m. As shown in Fig. 1a, its asymmetric unit cell contains one eight − coordinated Y3+ ion (denoted as Y1), one seven − coordinated Y3+ ion (denoted as Y2), one completely deprotonated BBTC4− ligand, five coordinated DMF molecules and two coordinated Cl− ions. Two crystallographically independent Y3+ ions show two different coordination geometries (Fig. 1b). The coordination geometry around Y1 ion can be described as a distorted dodecahedron, and its eight coordination oxygen atoms come from three BBTC4− ligands (O2, O3c, O4c and O8a) and four DMF molecules (O9, O9b, O10 and O10b). For Y2 ions, they are bonded with four oxygen (O1, O7a, O5d and O6d) from three BBTC4− ligands, the other oxygen (O12) from one DMF molecule and two Cl− ions (labeled Cl1 and Cl1b respectively), resulting in a distorted pentagonal bipyramid geometry. Among the four coordinated BBTC4− ligands, two carboxylate groups from two BBTC4− ligands coordinate with Y1 and Y2 centers in a chelating manner, while the other two carboxylate groups from two different BBTC4− ligands coordinate with Y1 and Y2 centers in a bidentate bridging manner, resulting in a binuclear Y2(CO2)2 subunit (SBU), in which the Y1⋅⋅⋅Y2 distance is 5.589 Å (Fig. 1c). The torsion angle of the diacetylene group (C9 − C10 − C11 − C12) is 0.994°, and the dihedral angle between two benzene rings is 0°. These data show that BBTC4− is nearly symmetrical, and the two benzene rings of BBTC4− are basically coplanar to meet the coordination geometry of Y3+. The Y–O bond lengths range from 2.227(7) to 2.397(6) Å (Table S1†), and the O − Y − O bond angles range from 51.8(2) to 161.0(2)° (Table S2†). These bond parameters are similar to the values of Re-carboxylate complexes in references [34, 35].

Fig. 1
figure 1

(a) Coordination environments of Y3+ ions with the H atoms omitted for clarity; Symmetry codes: a = x, y, 1 + z; b = x, 1.5 − y, z; c =  − 1 + x, y, z; d = 1 + x, y, 1 + z; b Coordination polyhedron of Y3+ ions; c One BBTC4− ligand links four Y2(CO2)2 SBUs; d One Y2(CO2)2 SBU is connected by four BBTC4− ligands; e Polyhedral representation of complex 1 seen from b direction (the green and turquoise polyhedra correspond to [Y1O8] and [Y2O5Cl2]units, respectively); f (4,4)-connected network presented, red represents the 4-connected node of organic ligand and blue represents the 4-connected node of U in complex 1 seen from b direction

Full size image

The carboxylate groups in complex 1 are coordinated with two Y3+ ions by chelating and bidentate bridging to form a binuclear Y cluster unit Y2(CO2)2 SBU. Generally speaking, each Y2(CO2)2 SBU is surrounded by four BBTC4− ligands, and each BBTC4− ligand connects four Y2(CO2)2 SBUs, resulting in a two-dimensional (2D) network framework (Fig. 1d and e). From the topological point of view, each Y2(CO2)2 SBU is regarded as a 4 − connected node and the organic ligand BBTC4− is considered as a 4 − connected node, the 2D structure of complex 1 can be described as a (4,4) − connected network (Fig. 1f).

3.2 Powder X-Ray Diffraction (PXRD) and Thermal Stability

Powder X–ray diffraction (PXRD) of complexes 1 and 2 were recorded at room temperature. As shown in Fig. S3†, the curves of the two complexes are similar, revealing that complexes 1 and 2 are isostructural with each other. The peak positions of the theoretical and experimental PXRD patterns are in good agreement with each other, which clearly indicate the high purity of these samples.

TGA measurements of complexes 1 and 2 were performed from room temperature to 700 °C under a nitrogen atmosphere (Fig. S4†), the isomorphism complexes 1 and 2 exhibit similar thermal behavior with several sequential weight loss processes. Considering the isomorphism of complexes 1 and 2, complex 1 is discussed as the representative example in detail. As shown in Fig. S4a†, Complex 1 shows sequential weight loss process with a percentage of 14.67% before 90 °C, corresponding to the loss of one guest H2O molecule and two guest DMF molecules per formula unit (calcd: 14.23%), and the second weight loss of 31.34% in 91 − 335 °C temperature range, attributing to the release of five coordinated DMF molecules per formula unit (calcd: 31.66%). Above 335 °C, the whole framework gradually begins to break down upon further heating.

3.3 Fluorescent Behavior

The solid − state luminescence properties of H4BBTC and complex 1 were investigated at room temperature. As shown in Fig. S5†, it is observed that H4BBTC exhibits a broad emission centered at 420 nm upon excitation at 340 nm due to the charge transfer in the organic ligand (π* → π) [36]. The electronic structure of Y3+ ion is a closed shell with no f–f transition, so it is a non-luminous rare earth ion. Therefore, at the same excitation wavelength, complex 1 displays luminescence with a single broad emission band centered at 402 nm, such a broad band emission of complex 1 can also be assigned to the ligand to ligand charge transfer of carboxylic acids [37]. In comparison to the emission of H4BBTC, the blue shift of the emission band (18 nm) may be attributed to the energy transfer from the ligand to the Y3+ ion [38,39,40].

The emission of complex 2 in the near–IR region was also investigated, which is a rarely described phenomenon [41, 42]. Three emission bands centered at 989 (5F5 → 5I7), 1177 (5F6 → 5I8) and 1478 nm (5F5 → 5I6) are observed for the Ho3+ complex (Fig. S6†).

In order to explore the luminescence properties of complex 1 in various organic solvents, emission spectra of complex 1 suspended in different solvents were studied. The details are as follows: the finely ground samples of complex 1 (2 mg) were dispersed into different solvents (2.0 mL), such as acetonitrile (MeCN), ethyl acetate, chloroform (CHCl3), dichloromethane (DCM), N, N-dimethylformamide (DMF), tetrahydrofuran (THF), benzene, toluene and acetone to form stable suspensions. As shown in Fig. 2, the emission intensities of complex 1 were dependent on the solvent molecules, and the results showed that complex 1 displayed the strongest luminescence emission intensity in MeCN solution. Undoubtedly, MeCN was selected as the solvent to carry out a series of luminescence sensing experiments.

Fig. 2
figure 2

Comparisons of the luminescence intensity of complex 1 in different solvents at room temperature

Full size image

3.4 Fluorescence Sensing Properties

3.4.1 Sensing of Cations

First, complex 1 (2 mg) was dispersed in the solution of M(NO3)x (2 mL) (Mx+  = Fe3+, Cu2+, Cr3+, Pb2+, Al3+, Ni2+, Co2+, Ca2+, Cd2+, Sr2+, Ba2+, Na+; 1 × 10−2 moL L−1). Then the fluorescence intensities of suspensions containing above 12 kinds of metal ions were recorded and compared, respectively. It can be clearly seen that among the all detected metal ions, the luminescence quenching ability of complex 1 to Fe3+ was the strongest, with a quenching efficiency of 90.26% for complex 1 (Fig. 3a). To further study, the luminescence titration experiments were performed by incremental addition of Fe3+ (1 mM) to the suspension of complex 1. It can be seen from Fig. 3b and c that the fluorescence intensity of complex 1 decreased gradually with the increasing concentration of Fe3+, and the fluorescence intensity of complex 1 was quenched to over 90% when the amount of Fe3+ ions was added to 60 μL. The quenching effect of Fe3+ ions on complex 1 can be explained quantitatively by the Stern–Volmer (SV) equation: (I0/I) = 1 + Ksv[C]. In the equation, I0 and I are the intensities of complex 1 before and after addition of Fe3+, [C] is the concentration of Fe3+ and Ksv is the quenching constant [43]. The Sterm-Volmer plot for Fe3+ is nearly linear at low concentrations and subsequently deviated from linearity, bending upwards at higher concentrations (Fig. 3d). The nonlinear nature of the SV plot can be ascribed to a self-absorption or energy transfer process [44,45,46].

Fig. 3
figure 3

a Luminescence quenching efficiency of complex 1 in various metal ions (1 × 10−2 mol L−1); b Emission spectra of complex 1 dispersed in MeCN after addition of Fe3+ solution (1 mM); c Quenching percentages of complex 1 upon addition of different volumes of 1 mM Fe3+ solution; d Stern–Volmer plots of complex 1 dispersed in MeCN at different concentrations of Fe3+ ions, the insert plot shows the correlation between the concentration of Fe3+ and (I0 − I)/I

Full size image

According to the equation calculation, the quenching constant of Fe3+ ion on complex 1 is calculated to be 1.45 × 105 M−1, which is larger than most of the reported MOF-based Fe3+ sensors, and the limit of detection (LOD) is calculated to be 0.12 μM, which is relatively low in comparison to that in the reported examples (Table S3†), indicating that at very low concentrations, complex 1 still has a significant fluorescence response to Fe3+, which shows that complex 1 has good detection sensitivity for Fe3+ [47]. To examine the sensing selectivity of complex 1 towards Fe3+ ions, the anti-interference experiments were carried out, and the results showed that the quenching efficiency of complex 1 toward Fe3+ was hardly affected by other interfering metal ions, indicating that Fe3+ had good anti-interference capability (Fig. 4a) [48]. Additionally, others fluorescence titration experiments were performed by adding other metal ions as well, as shown in Fig. 4b, the selectivity of complex 1 for Fe3+ was found to be very high by comparison with other ions [49].

Fig. 4
figure 4

a Luminescence intensities of complex 1 in different metal ions (1 × 10−2 mol L−1) and mixed-metal solutions containing Fe3+ (1 × 10−2 mol L−1); b Stern–Volmer plot of I0/I-1 versus different metal ions concentration in MeCN

Full size image

3.4.2 Sensing of NAEs

Similar to the fluorescence detection of metal ions, NAEs (4-nitrophenol (4-NP), 2,4,6-trinitrophenol (TNP) and 4-nitrotoluene (4-NT)) with a concentration of 5 × 10−3 moL L−1 were selected for sensing experiments. As shown in Fig. S7†, all the NAEs have a certain fluorescence weakening effect on complex 1, while TNP have excellent luminescent quenching efficiency, the quench order is as follows: TNP > 4-NP > 4-NT. This result indicate that complex 1 possesses a highly selectivity for sensing TNP compared to other nitro-aromatic explosives [50].

For further study the luminescent quenching efficiency, the titration experiments were performed to explore the sensitivity of complex 1 to three different NAEs, respectively. It was found that the fluorescence intensity of complex 1 gradually weakened with the increasing of NAEs concentrations, and the fluorescence quenching process can be observed in Fig. 5 and S8–S9†. Interestingly, the quenching efficiency can be as high as 94% after adding 100 μL TNP solution (Fig. 5a and b). In the low concentration range of NAEs, the calculated Ksv of complex 1 for the above three NAEs are shown in Table S4†, the quenching constant of complex 1 for TNP is calculated to be 4.08 × 104 M−1 (Fig. 5c), which is equivalent to some recent reports of MOF-based TNP sensors, and the LOD value for TNP is calculated to be 0.52 μM, which is comparable to most of reported MOF-based sensors (Table S5†). The LOD of TNP is small, indicating that complex 1 has high sensitivity for detecting TNP [51].

Fig. 5
figure 5

a The changes of emission spectra upon incremental addition of TNP solution (5 mM); b Quenching percentages of complex 1 upon addition of different volumes of 5 mM TNP solution; c S–V plot of complex 1 upon incremental addition of TNP solution (5 mM) in MeCN

Full size image

3.5 Recyclability and Stability of Complex 1

Further more, the recyclability experiments of complex 1 in sensing Fe3+ and NAEs processing were further explored, the suspensions of complex 1 were centrifuged and washed with CH3CN after every titration measurement cycle. As depicted in Fig. 6 and S10–S11†, it was found that complex 1 can still maintain good luminescence intensity and the quenching ability after being used for four times. The PXRD patterns reveal the positions and intensities of complex 1 have not significantly changed after four quenching recovery cycles, showing the good recyclability and high stability of complex 1 (Fig. S12†) [49].

Fig. 6
figure 6

The emission intensities of complex 1 for detecting Fe3+ (a) and TNP (b) in four fluorescence quenching-recovery cycles

Full size image

3.6 Fluorescence Quenching Mechanism of Complex 1

The fluorescent response rates of complex 1 to Fe3+ and TNP at different immersing time intervals show that the fluorescence intensity can reach the least values quickly within one minute (Fig. S13 and S14†), which can be deduced that the fluorescence quenching mechanism is not caused by the adsorption of Fe3+ or TNP in complex 1 [52]. In addition, the ultraviolet–visible absorption spectra of metal ions in MeCN solution (5 × 10–4 mol L−1) were measured, respectively. As shown in Fig. S15†, the absorption spectrum of Fe3+ overlaps with the emission spectrum of complex 1 to a certain extent, but there is less overlap with other metal ions. It means that Fe3+ can effectively absorb the energy of the emitted light of complex 1, and the luminescence quenching mechanism is mainly energy resonance transfer [53,54,55]. Subsequently, the ultraviolet–visible absorption spectra of NAEs (1 × 10–4 mol L−1) in MeCN solution were also tested. The results are shown in Fig. S16†, the ultraviolet absorption spectra of the three NAEs overlap with the emission spectra of complex 1 to a certain extent, but the spectral line of TNP overlap the most, indicating that the luminescence quenching of complex 1 is due to the NAEs effectively absorbing the energy of the emitted light of complex 1, and the quenching mechanism is mainly energy resonance transfer [56, 57].

4 Conclusions

In summary, two isomorphic rare earth MOFs (Re-BBTC) with 2D layers composed of binuclear secondary building units and BBTC4– were successfully synthesized under solvothermal conditions. Fluorescence experiments showed that Y-BBTC had efficient response in detecting Fe3+ and TNP with good sensitivity and selectivity. Combined with UV–Vis experimental results, the fluorescence quenching may be attributed to the resonance energy transfer between the analytes and the skeletons.