Highly Sensitive Fluorescence Quantitative Detection of Mercury in Soil Based on Non-labeled Molecular Beacon and Fluorescent Dye Hoechst 33258.

A highly sensitive and selective fluorescence method of quantitative detection for mercury in soil was developed using non-labeled molecular beacon (MB), single-stranded nucleic acid (ssDNA) and fluorescent dye Hoechst 33258. In this analytical method, the loop of MB was designed to be a sequence that was complementary to the ssDNA with multiple T-T mismatches, the stem of MB was completely designed as C-G base pairs, and both ends of the MB are not modified by any fluorophore and quencher. In the absence of Hg, the interaction between Hoechst 33258 and the MB was very weak, and the fluorescence signal of Hoechst 33258 was very low. In the presence of Hg, the MB and ssDNA with multiple T-T mismatches formed a double-stranded nucleic acid (dsDNA) via the T-Hg-T coordination structure which provided binding sites for Hoechst 33258. Then Hoechst 33258 binded to A-T base pairs of dsDNA, and the fluorescence intensity of Hoechst 33258 was significantly enhanced. Thus, a highly sensitive fluorescence quantitative detection method for Hg can be realized. In this strategy, the optimal determination conditions for Hg were a buffer solution pH of 8.2, an incubated temperature of 50°C, an incubated time of 5 min and NaCl of 60 mmol L. Under the optimum conditions, the fluorescence intensity of Hoechst 33258 exhibited a good linear dependence on the concentration of Hg in the range of 5 × 10 - 400 × 10 mol L. The fitted regression equation was ΔI = 2.1084C - 8.9587 with a correlation coefficient of 0.9943 (R), and the detection limit of this method was 3 × 10 mol L (3σ). The proposed method had a high selection; the common substances in soil such as Ca, Mg, Mn, Fe, Cu, Pb, Al, K, Na, Ni, Cd, Cr, SiO, Cl, PO, NH and S had no interference to the detection of mercury. The proposed method had a high accuracy, and it was applied to detect mercury of ten different types of soil; the recoveries were 97.65 - 103.22%. In addition, the proposed method had a low background emission, fast detection speed and low detection cost.