Construction Of The Ratiometric Fluorescence Sensor Based On Fe-MIL-88NH₂ And Egg White Au NCs And Its Application In The Rapid Detection Of Hg²⁺ And Thiram

Mercury ion(Hg²⁺),used in the manufacturing industry and in electrical products,is also a highly cytotoxic,corrosive and carcinogenic substance.Thiram,as a kind of high-efficiency pesticide,is commonly used to prevent and control the pests and diseases of vegetables,fruits,and food crops,and has been widely used in agricultural production.With the rapid development of science and technology,it has promoted the development of economic and improved people’s living standards.However,at the same time,frequent human activities cause Hg²⁺content and pesticide residues to increase year by year,exceeding the normal range,leading to deterioration of environmental quality and seriously endangering human health.Therefore,the rapid,accurate and sensitive sensing of Hg²⁺and thiram is particularly important.Ratiometric fluorescence sensors can improve sensitivity and accuracy by eliminating the effects of background fluorescence,light scattering from samples,and photobleaching of fluorophores through self-calibration of the two fluorescence peaks.Furthermore,a sensor of smartphone basd on ratiometric fluorescence can facilitate the visual monitoring of the target response through the high contrast between the fluorescence signals in two different bands.However,the research of object detection of deep learning,color feature recognition and linear analysis running simultaneously in one software is still lacking.In this study,a ratiometric fluorescence sensor based on Fe-MIL-88NH₂ and Au NCs synthesized with egg white as a template was constructed,and a deep learning-assisted smartphone detection platform was constructed to fast and accurate quantitative analysis of Hg²⁺and thiram.It has laid a foundation for effectively preventing heavy metal poisoning,preventing and controlling heavy metal and pesticide pollution,improving the quality and safety of agricultural products,and conducting on-site real-time detection under limited or non-concentrated resources.The main findings are as follows:（1）Fe-MIL-88NH₂ was synthesized by one-step hydrothermal method,and Au NCs were synthesized by reduction method with egg white as template.They were characterized by SEM,TEM,HRTEM,XRD,MALDI-TOF-MS,FTIR and fluorescence spectroscopy.It was found that Fe-MIL-88NH₂ has a spindle structure with a length and width of 1.1～1.3μm and 0.3～0.5μm,respectively,and has-NH₂ functional group,blue fluorescence emission at 427 nm,excitation-independent behavior.On the other hand,Au NCs has a“pine leaf”-like structure,an average particle size of 25 nm,a large number of-COOH functional groups,strong red fluorescence emission at 643 nm,and excitation-dependent behavior.Based on this,a ratiometric fluorescence sensing system（Fe-MIL-88NH₂/Au NCs）was constructed,in which Fe-MIL-88NH₂ served as the built-in signal at 427 nm with blue fluorescence,Au NCs,as a response signal,have strong red fluorescence at 643 nm.（2）The system of Fe-MIL-88NH₂/Au NCs can detect Hg²⁺and thiram by“on-off”and“on-off-on”.With the increase of Hg²⁺concentration,the fluorescence color of the system changed from pink to purple and finally to blue.The concentration of Hg²⁺in the range of0.012～10μmol/L and 10～21μmol/L showed good linearity with F₄₂₇/F₆₄₃ values,and the detection limit was 7 nmol/L.With the increase of the concentration of thiram,the fluorescence color of the system gradually returned to pink,and the F₄₂₇/F₆₄₃ values in the range of 0.083～12.477μmol/L and 12.477～41.590μmol/L of thiram showed a linear relationship.The detection limit was 0.083μmol/L.The recovery rate of Hg²⁺detected by this method in the actual sample was 98.21～100.81%,RSD≤1.8%,and the recovery rate of thiram was 99.64～103.20%,RSD≤1.88%.It has clear advantages over previously reported methods.（3）It was proved that the quenching mechanism of the system is dynamic quenching by using instruments such as ultraviolet-visible spectrophotometer,fluorescence spectrophotometer,and time-resolved fluorescence spectrometer.The recovery of the system’s fluorescence is due to the-SH group in thiram forming“Hg-S”with Hg²⁺to compete with Au NCs,inhibiting the dynamic quenching effect.In addition,through DFT calculation,there is electron transfer between Au NCs and Hg²⁺,which further proves that the quenching mechanism of the system is dynamic quenching.（4）A miniaturized imaging module that can simulate a fluorescence spectrophotometer is printed by 3D printing technology.At the same time,a We Chat applet with the Yolov3algorithm of deep learning was developed,which can realize automatic target detection,color feature extraction and linear analysis.The system of Fe-MIL-88NH₂/Au NCs,smartphone and imaging module are integrated together to build a smartphone sensing platform.The imaging module and the smartphone’s camera are used as detectors to record the emitted fluorescence intensity in the deep learning-assisted smartphone sensing platform.It is analyzed by the We Chat applet according to the red（R）,green（G）and blue（B）and hue（H）,saturation（S）and lightness（V）values of the image.According to the extracted R,G,B,H,S,V values,R/B and in a wide range of 0.002～30μmol/L of Hg²⁺have a good linear relationship.There was a linear relationship with R/G and thiram in a wide range of0.083～49.910μmol/L.The recovery rate of Hg²⁺detected by this platform in actual samples was 97.9～112.3%,RSD≤5.25%,and the recovery rate of thiram detection was92.01～102.82%,RSD≤4.86%.Compared with the previously reported platforms,the platform is portable,fast,sensitive,and accurate in detection.