A new luminescent metal-organic framework with 2,6-di(1H-imidazol-1-yl)naphthalene and biphenyl-3,4′,5-tricarboxylic acid

2.1 Materials and methods

All commercially available chemicals and solvents were of reagent grade and were used as received without further purification. Ligand L was prepared according to the reported procedures [16]. The C, H, and N contents were performed using an Elementar Vario MICRO elemental analyzer. Fourier Transform infrared spectroscopy (FT-IR) spectra were recorded in the range of 4000–400 cm⁻¹ on a Bruker Vector 22 FT-IR spectrometer using KBr pellets. Powder X-ray diffraction (PXRD) measurements were performed on a Bruker D8 Advance. Single crystal data was collected on a Bruker Smart Apex II CCD area-detector diffractometer (vide infra). Thermogravimetric analyses (TGA) were taken on a MettlerToledo (TGA/DSC1) thermal analyzer under nitrogen at a heating rate of 10 K min⁻¹. Photoluminescence spectra were obtained on an Aminco Bowman Series 2 spectrofluorometer with a xenon arc lamp as light source. In the measurements of emission and excitation spectra, the pass width is 5 nm.

2.2 Preparation of [Co(L)(Hbptc)(H₂O)] (1)

A mixture of L (13.8 mg, 0.05 mmol), Co(NO₃)₂·6H₂O (29.1 mg, 0.10 mmol), H₃bptc (14.3 mg, 0.05 mmol) in N,N-dimethylformamide (DMF)/H₂O (10 mL, v/v, 1:1) was sealed in a Teflon-lined stainless steel container and heated at 100 °C for 3 days. After being cooled to room temperature, the sample was obtained in 69% yield based on L. – Anal. for C₃₁H₂₂CoN₄O₇: calcd. C 59.91, H 3.57, N 9.02; found C 60.79, H 3.62, N 9.25%. – IR (KBr): ν (cm⁻¹) = 3151, 1708, 1605, 1540, 1509, 1466, 1299, 1114, 1068, 941, 844, 761, 691, 656, 466 (Figure 1).

2.3 X-ray crystallography

Crystallographic data for 1 was collected on a Bruker Smart Apex II CCD area-detector diffractometer with graphite-monochromated Mo Kα radiation (λ = 0.71,073 Å). The integration of the diffraction data as well as the intensity corrections for the Lorentz and polarization effects was carried out using the Saint program [17]. Semi-empirical absorption corrections were performed using Sadabs program [18]. The structure of 1 was solved by Direct Methods using Shelxs-2014 [19] and all the non-hydrogen atoms were refined anisotropically on F ² by the full-matrix least-squares technique with Shelxl-2014 [20]. The hydrogen atoms in the organic ligands were generated geometrically and refined isotropically using the riding model, while the ones of coordinated water molecule in 1 were found from the Fourier map directly. The details of the crystal parameters, data collection and refinements for the complexes are summarized in Table 1, and selected bond lengths and angles are listed in Table 2.

CCDC 2102760 for 1 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

3.1 Crystal structure description of [Co(L)(Hbptc)(H₂O)] (1)

Single crystal X-ray diffraction analysis revealed that complex 1 crystallizes in the monoclinic system, space group Cc, the asymmetric unit of 1 consists of one Co(II), one L, one Hbptc²⁻ and one coordinated water molecule. As exhibited in Figure 2a, Co1 is hexa-coordinated with distorted octahedral coordination geometry, surrounded by two nitrogen atoms (N1, N4A) from different L units, and four carboxylate oxygen atoms [O1, O4B, O3B, O7], three oxygen atoms come from two different Hbptc²⁻ [O1, O4B, O3B], and one from the coordinated water molecule (O7). The length of the Co–N bonds are 2.081(3) and 2.144(3) Å, while the Co–O bond ranges from 2.054(3) to 2.284(3) Å. The coordination angle around the Co(II) center in 1 are from 58.66(9) to 175.88(12)° (Table 2). Every L connects two Co(II) to form a chain (Figure 2b), and each Hbptc²⁻ connects two Co(II) using its two carboxylate groups with a (μ ₁-η ¹:η ⁰)-(μ ₂-η ¹:η ¹)-Hbptc²⁻ coordination mode (Scheme 1) to give another chain (Figure 2c), which are further linked by winding interactions to give the three-dimensional network structure (Figure 2d).

Figure 1:

Fourier Transform infrared spectroscopy (FT-IR) spectrum of 1.

Scheme 1:

Coordination modes of Hbptc²⁻ in 1.

3.2 Powder X-ray diffraction (PXRD) and thermal stability

The purity of the as-synthesized bulky sample was studied by PXRD measurements and the results are provided in Figure 3. The PXRD pattern of the as-synthesized sample is in reasonable agreement with the corresponding simulated one, implying a high content of 1 in the polycrystalline sample. TGA were performed to check the thermal stability of the frameworks, and the results are shown in Figure 4. In the case of 1, a weight loss of 2.43% was found in the temperature range of 290–320 °C, which is ascribed to the removal of water molecules (calcd 2.90%), and a further weight loss starts from about 340 °C on, corresponding to the collapse of the framework.

Figure 2:

(a) Coordination environment of Co(II) in 1 with ellipsoids drawn at 50% probability level. The hydrogen atoms are omitted for clarity. Symmetry codes: ^#1 x + 1/2, y − 1/2, z + 1; ^#2 x + 1/2, −y + 1/2, z + 1/2. (b) Chains of the Co(II)-L ligand. (c) Chains of the Co(II)-Hbptc²⁻ ligand. (d) 3D framework of 1.

Figure 3:

Experimental and calculated PXRD patterns of 1.

Figure 4:

TG curve of 1.

3.3 Photoluminescence of 1

We have studied the photoluminescence emission spectra of 1 in solid state at room temperature. As illustrated in Figure 5, 1 shows a strong emission peak at 361 nm (λ _ex = 290 nm). The emissions of 1 may be caused by π* − π transitions in the ligand and the charge transfer transition between L and Co(II) [11, 21, 22].

Figure 5:

Solid-state photoluminescence spectra of 1 at room temperature.

3.4 Photoluminescence quenching-based sensing of Fe³⁺ ions

1 was investigated with respect to its selective recognition ability. The as-synthesized sample of 1 (2 mg) was immersed in DMF solutions containing different salts of the class M(NO₃)_x (2 mL), [M] = 10⁻³ mol/L (M = Li⁺, Na⁺, K⁺, Zn²⁺, Mn²⁺, Al³⁺, Cu²⁺, Cd²⁺, Co²⁺, Ni²⁺ and Fe³⁺). All metal ions affect the luminescence intensity of 1, but there is no significant drop for most species (Figure 6a and b). Compared to other cations, however, when adding Fe³⁺ ions to a suspension of 1, it cause a drastic fluorescence quenching, suggesting that 1 has a hugely selectivity for Fe³⁺ ions sensing.

Figure 6:

(a, b) Photoluminescence intensities of 1 dispersed in DMF with variable concentrations metal ions when excited at 290 nm.

Moreover, titration experiments were carried out to study the luminescence response of 1 to Fe³⁺ ions. Photoluminescence titration was used to detect the response of varying concentrations of Fe³⁺ ions to 1 dispersed in DMF. A progressive quenching occurred with the addition of Fe³⁺ ions from 0 to 100 μL (Figure 7a), the quenching efficiency is almost as high as 90%. Furthermore, quenching efficiency can be quantitatively analyzed by the Stern-Volmer (SV) equation:

Figure 7:

(a) Photoluminescence quenching plots of 1 in DMF in the presence of different Fe³⁺ concentration (excited in 290 nm). (b) Stern-Volmer linear plot for 1 in the presence of Fe³⁺ ions.

I ₀ and I are respectively the luminescence intensity of 1 dispersed in DMF and upon addition of Fe³⁺ ions, K _sv is the quenching constant, [M] is the molar concentration of Fe³⁺ ions. Take the luminescence intensity of the top six groups (Figure 7a). The slope of this linear fitting equation is 2.40 × 10⁴, so the resulting K _sv was calculated as 2.40 × 10⁴ M⁻¹ (Figure 7b). According to the formula of 3σ/k (σ: standard error; k: slope), the limit of detection (LOD) was calculated to as 209 ppb. This is equivalent to the observation on previously reported MOFs able to sense Fe³⁺ ions [23], [24], [25], which confirms the sensibility of 1.

According to reports, Fe³⁺ ions exhibit a broad absorption band in the UV–vis absorption spectrum between 250 and 400 nm. According to the photoluminescence emission spectrum of 1 ( Figure 5), the emission band of 1 is between 325 and 427 nm. There is a large overlap between the peaks, whereas other cations have no obvious absorption bands in this range. Therefore, it can be suggested that the fluorescence quenching of 1 may be caused by the competition between Fe³⁺ ions and 1 [22, 24, 26].