Unless otherwise stated, starting materials were commercially available and analytically pure, and the solvent was dried before use. The water used was redistilled water. The UV absorption and fluorescence emission spectra were recorded on a U-2550 Double-beam UV-Vis spectrophotometer (Japan) and a F-7000 fluorescence spectrometer (Japan), respectively. Melting points were recorded using an uncorrected X-4digital melting point apparatus. NMR were recorded on a Bruker Avance 400 MHz spectrometer (CDCl₃ and DMSO-d ₆ ) with resonances relative to tetramethyl-silane (TMS) as an internal standard. Mass spectra (ESI) were recorded on a Waters XEVO G2-XSmass spectrometer. Fluorescence images of cells were analyzed by Dual photon confocal microscope (Olympus FV1000 MPE).

FPS was formulated into 1.0 mmol/L solution in DMSO and then diluted to definite concentration with methanol before the spectral experiment. The salts used in standard stock solutions of metal ions were MnSO₄, Ca(NO₃)₂, CuSO₄, Al(NO₃)₃, Ba(NO₃)₂, AgNO₃, CoCl₂, ZnSO₄, FeSO₄, NaCl, KCl, Pb(NO₃)₂, MgSO₄, Fe(NH₄) (SO₄)₂ in distilled water to prepare 0.050 mol/L. The spectral changes of the mixed solutions of FPS with various metal ions were studied by UV-Vis absorption and fluorescence spectroscopy at room temperature. The fluorescence emission of FPS were recorded with excitation at 480 nm.

The cytotoxicity of FPS was researched by CCK8 assay according to reported methods (Hou et al., 2019). Hela cells were cultured in DMEM medium containing 10% fetal bovine serum cell culture medium with 5% CO₂ atmosphere at 37°C. The cells were transferred into 24-well plates and incubated for 24 h at 37°C. FPS, diluted to the desired concentrations (50–1,000 μmol/L) in culture medium, was added to the well. Then the original medium was removed after 24 h, and 10 μl of CCK-8 solution (5 mg/ml stock) was added to the Well and incubated for 2 h at 37°C. Absorbance at 450 nm was recorded with an enzyme-linked immunosorbent assay (ELISA) reader (Bio-Tek). The results showed that FPS exhibited low cytotoxicity against Hela cell lines with IC₅₀ more than 500 μmol/L.

FPS showed low fluorescence emission could be because of the photoinduced electron transfer (PET) process between the imine group and the benzene ring. To research the diverse solvent effect on FPS and FPS-Zn ²⁺ , the emission and excitation spectra of FPS and FPS-Zn ²⁺ were recorded in different solvents [MeOH, EtOH, DMSO, and DMF (Figure 1)], respectively. These results showed that FPS and FPS-Zn ²⁺ have higher fluorescence enhancement in MeOH than the other solvents.

The fluorescence properties of FPS with various metal ions (10.0 equiv. of Na⁺, K⁺, Ba²⁺, Zn²⁺, Mn²⁺, Mg²⁺, Ca²⁺, Ag⁺, Co²⁺, Ni⁺, Fe³⁺, Cd²⁺, Pb²⁺, Cu²⁺, Al³⁺) were investigated, respectively. Results showed that Zn²⁺ caused significant fluorescent enhancement at 480 nm with the color change from colorless to the light blue. These phenomena indicated that FPS could be used as fluorescent sensors for Zn²⁺ recognition (Figure 2A). The ability was also explored of FPS to detect Zn²⁺ in the presence of other metal ions. A competitive test was carried out, in which 10 equivalent other metal ions were added to the solution of FPS and Zn²⁺ ion (Figure 2B), respectively. Figure 2B shows that the fluorescence quenching of FPS-Zn²⁺ undergoes a significant change with the addition of Cu²⁺ and Fe²⁺, and there is a little interference with the addition of Al³⁺, Fe³⁺, Co²⁺ and Ni²⁺ ions.

Figure 3A is the Job’s plot of the fluorescence signal for FPS and Zn²⁺ solutions. The binding stoichiometry can be obtained from the plot. It revealed that a 1:1 binding was obtained between FPS and Zn²⁺ in methanol solution. Then the fluorescence characteristics of FPS to Zn²⁺ were further studied by fluorescence titration experiments (Figure 3B). The stability constant of FPS and Zn²⁺ was calculated to be 4.45×10⁴ (r ² = 0.9626) from the nonlinear least squares fitting of the data, according to the Benesi-Hildebrand equation (Scheme 2). As shown in Figure 3B, with gradual addition of 0–20 μmol/L Zn²⁺ into the methanol solution of FPS (20.0 μmol/L), the fluorescence emission at 480 nm was increased gradually. Moreover, the detection limit of FPS to Zn²⁺ was calculated (LOD = 3σ/slope) to be 1.19 × 10^–8 mol/L (r ² = 0.9959).

Based on the results in Figure 2B, to evaluate further the effect of Cu²⁺ and Fe²⁺ concentration on the probe FPS for recognition Zn²⁺, respectively, the fluorescence properties of complex FPS-Zn²⁺ were studied in methanol solution (Figure 4). For Cu²⁺ ion, at low concentrations (c_Cu2+ ≤ 5 μmol/L), the fluorescence emission was continuously quenched with the increase of Cu²⁺ and there was a good linear relationship. However, for Fe²⁺, there were no similar phenomena, the fluorescence emission at 480 nm was quenched completely, at that moment, with the addition of Fe²⁺ even in small amounts. Then to investigate the time-dependent of fluorescence quenching for Cu²⁺, as shown in Figure 4, the florescence intensity tended to be stable after adding the Cu²⁺ ion 15 seconds. In addition, the detection limit of complex FPS -Zn²⁺ to Cu²⁺ was calculated (LOD = 3σ/slope) to be 2.25 × 10⁻⁷ mol/L (r ² = 0.9987). These results show that a fluorescent probe composed of complex FPS -Zn²⁺ could be used as fluorescent sensors for Cu²⁺ detection real-time and for Fe²⁺ qualitative determination.

As shown in Figure 5 , the fluorescent spectra of FPS (5 μmol/L) + 1.0 equiv. Zn²⁺ with pH (1–12) in DMSO/H₂O be studied, in the neutral solution, the fluorescence intensity of FPS + Zn²⁺ complex is strongest. At the same time, when the pH lower than five or higher than 10, the fluorescent of FPS + Zn²⁺ complex is quenching.

The binding ability of FPS with Zn²⁺ was evaluated using ¹H NMR. As shown in Figure 6, when Zn²⁺ was added to FPS, the protons on the phenolic hydroxyl and amide of the FPS Ha and Hb were nearly despaired, respectively. This indicated that phenolic hydroxyl and amide coordinated to Zn²⁺ and the FPS-Zn²⁺ complex was formed.

Based on the experimental data and Job’s plot, to further elucidate the influence of the structure on the electronic properties, DFT calculations are performed for FPS, FPS-X (X = Zn²⁺, Fe²⁺, Cu²⁺). As shown in Figure 7, for complex FPS-X (X = Zn²⁺, Fe²⁺, Cu²⁺), the deprotonated O atom of phenolic hydroxyl, O atoms of carbonyl, imine and the O atom of methanol were coordinated with metal ions. The calculated distributions of molecular orbitals (HOMO, highest occupied molecular orbital; SOMO, single electron occupied molecular orbital; and LUMO, lowest unoccupied molecular orbital) are shown in Figure 7. The HOMO and LUMO of FPS are predominately determined by the phenol moiety, the bride including acyl hydrazine moiety, furan moiety, and sectionally listed in pyridine moiety, respectively. Once FPS coordinated to metal ions (Zn²⁺, Fe²⁺, Cu²⁺), the HOMO/SOMO and LUMO in FPS-X (X = Zn²⁺, Fe²⁺, Cu²⁺) were conversely localized on the pyridine moiety and metal ions one, respectively. Apparently, this phenomenon was attributed to the enlargement of the conjugated system due to the complexation of the FPS and metal ions. The lower energy gap between the HOMO and LUMO level of FPS-X (X = Zn²⁺, Fe²⁺, Cu²⁺) compared with 4.21 eV of FPS was in good agreement with the red shift of the experimental fluorescence spectra (Figure 2B). Furthermore, the calculational binding energy E _bind (Supplementary Table S1) of FPS-X (X = Zn²⁺,Fe²⁺, Cu²⁺) show that the minimal values of FPS-Zn²⁺ and the maximum values of FPS-Cu²⁺ which cause the greater complexation of Cu²⁺compared with Zn²⁺ and Fe²⁺, probably further leading to the fluorescence quenching of FPS-Cu²⁺.

Fluorescence calculations of FPS-X (X = Zn²⁺, Fe²⁺, Cu²⁺) (Supplementary Tables S2–S4) was in good agreement with the experimental fluorescence data.

Hela cells were cultured in DMEM medium containing 10% fetal bovine serum cell culture medium with 5% CO₂ atmosphere at 37°C. The cells were transferred into 24-well plates and incubated for 24 h at 37°C. The first group Hela cells treated in a culture medium (DMSO: DMEM = 1: 99, v/v) alone were used as a control (Figure 8A). In group 2 and 3, Hela cells were cultured with probe FPS solution (5.0 μmol/L) for 25 min (Figure 8B). In groups 4 and 5, Hela cells were cultivated successively with probe FPS (5.0 μmol/L and 10.0 μmol/L) for 25 min, after being washed with PBS, and further incubated with Zn²⁺ (10.0 μmol/L and 20.0 μmol/L) for 5 min, respectively (Figure 8C).