Introduction

The design and development of fluorescent probes for the detection of heavy and transition metals are significant due to their vital role in biological and environmental application in recent years [1–4]. Mercury is one of the most prevalent toxic metals in both the environment and biological system [5, 6]. Even at a very low concentration, the mercuric ion (Hg2+) which combines with both inorganic and organic ligands, can readily penetrate through biological membranes. Mercury can cause serious and irreversible DNA damage, mitosis impairment and nervous system defects [7–9]. Therefore, it is of great importance to develop advanced methods for detecting mercury ions in biological system and natural environment. Until recently, many excellent works of Hg2+ sensing by synthesized fluorescent probes have been reported and investigated [10–12].

Among these works been reported, many fluorescent probes were based on single emission intensity change [13–15]. However, changes in the emission intensity at a single wave length being the only detection signal, such turn-on probes tend to be affected by the variations in the sample and probe environment, illumination intensity or instrumental efficiency [16]. Ratiometric probes can eliminate most of these interferences through simultaneous recording ratio signals of two emission intensities at different wavelengths, which provided a built-in correction for the environmental effects [17, 18]. From this point of view, 2-(2′-hydroxyphenyl) benzothiazole (HBT) is very familiar because of its intramolecularly hydrogen-bonded property, which exhibits excited state intramolecular proton transfer (ESIPT) [19, 20]. There were a number of reactive probes reported based on the HBT moiety for the selective detection of different analytes via “protection-deprotection” sequence [21–23]. According to the strong thiophilic affinity of Hg2+, many chemodosimeters contained an “S″ group [7, 24, 25]. Thus, along with the leaving of HgS, the “protection-deprotection” reaction is accomplished, as well as the resulting ESIPT modulated fluorescence off–on response.

Herein, we present a simple and new fluorescent probe BTP for the detection of Hg2+ based on ESIPT mechanism. Probe BTP contained an “P = S″ group, and as expected, it exhibited a nonreversible, highly selective and sensitive recognition toward Hg2+ over other examined metal ions in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Additionally, according to the fluorescence imaging experiments of Hg2+ ions in living TE-1 cells, BTP could be used for detecting Hg2+ in biological samples.

Experimental Section

Apparatus

Fluorescence spectra were recorded on the F-7000 FL Spectrophotometer (Hitachi, Japan), and the excitation and emission wavelength band passes were both set at 5.0 nm. 1H and 13C NMR spectra were recorded using a Bruker DTX-400 spectrometer. Samples were dissolved in CDCl3 and placed in 5 mm NMR tubes, TMS was used as internal reference. ESI mass spectra were carried out on an HPLC Q-Exactive HR-MS spectrometer (Thermo, USA) by using methanol as mobile phase. Fluorescence images experiments were carried out with a Zeiss-Axio Observer D1 inverted fluorescence microscope.

Materials

All chemicals reagents were used as received from commercial sources without further purification. Solvents for chemical synthesis and analysis were purified according to standard procedures. Deionized water was used throughout the experiment. Chloride salts of metal ions (Li+, K+, Na+, Ca2+, Mg2+, Ba2+, Zn2+, Fe2+, Mn2+, Cu2+, Co2+, Ni2+, Cd2+, Cr3+, Hg2+, Al3+) and the nitrate salts of Ag+, Pb2+ and Fe3+ ions were prepared as 10.00 mM in water solution.

Synthesis of Probe BTP

The synthetic routine of probe BTP is outlined in Scheme 1. HBT was synthesized by a similar way described in a reported method [26]. Dimethylthiophosphinoyl chloride (105 μL, 1 mmol) in 20 mL dichloromethane was added to a mixture of HBT (136 mg, 0.60 mmol) and triethylamine (138 μL, 1 mmol) in 20 mL dichloromethane. The reaction mixture was stirred at room temperature for 4 h, and concentrated under reduced pressure. The crude product was purified by column chromatography (silica gel, hexane/ethyl acetate =2/1) to afford BTP (124 mg, 65 %) as a white powder. 31P NMR (162 MHz, CDCl3): δ = 96.16 ppm. 1H NMR (400 MHz, CDCl3, ppm) δ: 2.11 (d, 3 H, J = 4 Hz), 2.13 (d, 3 H, J = 4 Hz),7.35 (t, 1 H, J = 8 Hz), 7.50 (m, 3 H), 7.88 (d, 1 H, J = 8 Hz), 7.96 (d, 1 H, J = 8 Hz), 8.13 (d, 1 H, J = 8 Hz), 8.33 (d, 1 H, J = 4 Hz); 13C NMR (100 MHz, CDCl3, ppm) δ: 23.88, 24.61, 121.40, 121.69, 121.74, 123.31, 125.21, 125.38, 125.84, 125.89, 126.41, 130.81, 131.47, 135.54, 148.80, 148.89, 152.63, 162.45; HR-MS m/z: Calcd for C15H15NOPS2 + ([M + H+]+) 320.0333, found 320.0315 [M + H+]+, 342.0129 [M + Na+]+.

Scheme 1
scheme 1

Synthetic route of probe BTP

Full size image

Results and Analysis

Probe BTP was dissolved in CH3CN to make a 1 mM stock solution. Then the stock solution was further diluted to require concentration for measurement.

Fluorescence Spectral Responses of BTP

As is well known, the HBT and its derivatives produced the ESIPT tautomers (the keto forms), which showed fluorescence more powerfully at longer wavelengths compared to the phenol forms upon irradiation. The selectivity of BTP was observed in the fluorescence emission profile of BTP (10 μM) in a CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution with appropriate amounts of metal ions (Fig. 1. inset). Probe BTP alone displayed an emission band centered at 377 nm, when excited at 310 nm. Upon addition of 10 eq. Hg2+, the emission at 377 nm decreased, and a significant enhancement at 470 nm emerged quickly. This indicated that the chemical reaction between Hg2+ and the receptor (thiophosphinated phenolic) started at this minimum concentration and thus the ESIPT properties of HBT were demasked (Scheme 2). The fluorescence intensity ratio (F470/F377) of probe BTP toward different metal ions was recorded in Fig. 1, which exhibited a prominent enhancement of the fluorescence ratio (F470/F377) in the presence of 10 eq. Hg2+. In the meantime, no response could be observed upon the addition of the same amount of other ions. This strongly suggested that BTP can serve as a high sensitivity for Hg2+.

Fig. 1
figure 1

Fluorescence intensity ratio (F470/F377) of BTP (10 μM) in the presence of 10 eq. different metal ions in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Inset: Fluorescence spectra of BTP (10 μM) in the presence of 10 eq. different metal ions in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Λex = 310 nm, scan range 330–600 nm, slit width 5 nm

Full size image
Scheme 2
scheme 2

Hg2+-promoted deprotection of BTP to compound HBT

Full size image

Furthermore, as shown in Fig. 2, the ratiometric fluorescence signal response of probe BTP toward Hg2+ in the presence of various coexistent anions such as NO3 −, NO2 −, Cl−, Br−, PO4 3−, SO4 2− and CO3 2−, which revealed that all the tested anions have little interference on the detecting of Hg2+. It was also investigated that the competitive experiment also confirmed that the background metal ions showed very low interference with the detection of Hg2+ (Fig. S5), only Cu2+ has posed a negligible effect on the fluorescence response of BTP for Hg2+, it may due to the quenching effect of the paramagnetic Cu2+ [27]. Therefore, these results suggested that probe BTP has a high selectivity for Hg2+ in the presence of these tested foreign metal ions and anions.

Fig. 2
figure 2

Fluorescence intensity ratio (F470/F377) of BTP (10 μM) upon addition of 10 eq. Hg2+ in the presence of 10 eq. background various coexistent anions in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Λex = 310 nm, scan range 330–600 nm, slit width: 5 nm

Full size image

As shown in Fig. 3, the time dependence of the response of BTP to Hg2+ ions was investigated. It can be seen that the fluorescence intensity ratio signal of the BTP with Hg2+ ion increased for a few seconds, and leveled off as the time continues, while the fluorescence intensity of blank solution (only BTP, 10 μM) showed almost unchanged at the same conditions. The time-dependent change plot demonstrated the reaction could complete in about 1 min, which indicated the probe BTP had a fast response for Hg2+. Therefore, a 1 min reaction time was selected in subsequent experiments in order to make the metal ions chelate with the sensors sufficiently.

Fig. 3
figure 3

Effect of reaction time on fluorescence intensity ratio (F470/F377) of BTP (10 μM) in the absence and presence of 10 eq. Hg2+ in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Λex = 310 nm, scan range 330–600 nm, slit width 5 nm

Full size image

To further investigate the interaction between Hg2+ and probe BTP, a fluorescence titration experiment was carried out. The fluorescence spectra of BTP (10 μM) exposed to CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution containing different concentrations of Hg2+ were then recorded at an excitation wavelength of 310 nm (Fig. 4). Upon treatment with increasing concentrations of Hg2+, the fluorescence intensity ratio (F470/F377) gradually increased, and reached saturation when the amount of Hg2+ was more than 12 μM (Fig. S6). Moreover, a linear relationship was found between the fluorescence intensity ratio (F470/F377) and the Hg2+ concentration from 4 to 12 μM (Fig. S7), the detection limit (3σ/slope) of probe BTP for the determination of Hg2+ was found to be 12 nM [28, 29]. These results demonstrated that probe BTP could detect Hg2+ quantitatively.

Fig. 4
figure 4

Fluorescence emission spectra of BTP (10 μM) with gradual addition of various amounts of Hg2+ (from bottom 0–2 eq.) in CH3CN/HEPES (10 mM, pH = 7.4, 1:4, v/v) solution. Λex = 310 nm, scan range 330–600 nm, slit width: 5 nm

Full size image

For practical applicability, the proper pH condition of this new probe for Hg2+ detection was also evaluated. We investigated the fluorescence properties of probe BTP and that probe BTP with Hg2+ (10 eq.) under different pH values, respectively (Fig. 5). Probe BTP was pH insensitive, and its ratiometric fluorescence response (F470/F377) was quite weak from pH 4 to 9. However, the fluorescent sensing toward Hg2+ was obviously affected by the change of pH values. Ratiometric fluorescence response (F470/F377) reached its maximum and kept constant around biologically relevant pH 4 to 9, indicating that its potential for application in biological systems.

Fig. 5
figure 5

Ratiometric fluorescence response (F470/F377) of free BTP (10 μM) and in the presence of 10 eq. Hg2+ in CH3CN/HEPES (10 mM, 1:4, v/v) solution with different pH (10 mM HEPES) conditions. Λex = 310 nm, scan range 330–600 nm, slit width: 5 nm

Full size image

Mechanism

In addition, the KI-adding experiments were conducted to examine the reversibility of this reaction as shown in Fig. S8. When excess KI (2 eq. of Hg2+) was added to the BTP (10 μM) and Hg2+ (100 μM) in CH3CN/HEPES (10 mM, 1:4, v/v) solution, the fluorescence intensity at 470 nm almost unchanged, indicating that the coordination of BTP with Hg2+ was chemically nonreversible.

According to the previous reported work [30–32], we proposed that the strong fluorescence enhancement was attributed to the deprotection of dimethyl- thiophosphinoyl group and concurrent generation of HBT. The observed change in presence of Hg2+ may arise from the HBT moiety which was released from BTP with the leaving of HgS. In order to verify this speculation, the reaction products of probe BTP and Hg2+ were subjected to ESI-HRMS spectrum. A major ion peak was founded at m/z = 228.0469 (Fig. S9), corresponding to the resulting HBT ([M + H]+), clearly confirmed the proposed mechanism as shown in Scheme 2.

Bioimaging Applications of Probe BTP in TE-1 Cells

We further investigated the practical application of BTP in biological systems [33, 34]. Fluorescent imaging inside TE-1 cells was monitored by fluorescence microscopy. As shown in Fig. 6, very weak fluorescence of BTP inside the living TE-1 cells was observed (Fig. 6b). After washing with water twice, 30 μM of Hg2+ were then supplemented to the cells. After incubated at 37 °C for another 30 min, a significant increase in the fluorescence from the intracellular area was observed (Fig. 6d). A bright field transmission image of cells with BTP and BTP with Hg2+ confirmed that the cells were viable throughout the imaging experiments (Fig. 6a and c). Therefore, these results demonstrated that probe BTP was cell membrane permeable and capable of fluorescence imaging of Hg2+ in biological samples.

Fig. 6
figure 6

Fluorescence images of Hg2+ in TE-1 cells with 10 μM solution of BTP in PBS buffer for 30 min at 37 °C, bright-field transmission images (a, c) and fluorescence images (b, d) of TE-1 cells incubated with 0 μM, 30 μM of Hg2+ for 30 min, respectively (λex = 340 nm, blue channel)

Full size image

Conclusion

In summary, a new ESIPT-based ratiometric fluorescence probe BTP for Hg2+ was prepared and reported. Probe BTP exhibited highly selective binding with Hg2+ over other metal ions and various coexistent anions in CH3CN/HEPES (10 mM, 1:4, v/v) solution. Moreover, the preliminary experimental results demonstrated that probe BTP could be used for detecting Hg2+ in biological samples.