Machine Learning System To Monitor Hg²⁺ and Sulfide Using a Polychromatic Fluorescence-Colorimetric Paper Sensor

Abstract

An optical monitoring device combining a smartphone with a polychromatic ratiometric fluorescence-colorimetric paper sensor was developed to detect Hg2+ and S2– in water and seafood. This monitoring included the detection of food deterioration and was made possible by processing the sensing data with a machine learning algorithm. The polychromatic fluorescence sensor was composed of blue fluorescent carbon quantum dots (CDs) (BU-CDs) and green and red fluorescent CdZnTe quantum dots (QDs) (named GN-QDs and RD-QDs, respectively). The experimental results and density functional theory (DFT) prove that the incorporation of Zn can improve the stability and quantum yield of CdZnTe QDs. According to the dynamic and static quenching mechanisms, GN-QDs and RD-QDs were quenched by Hg2+ and sulfide, respectively, but BU-CDs were not sensitive to them. The system colors change from green to red to blue as the concentration of the two detectors rises, and the limits of detection (LOD) were 0.002 and 1.488 μM, respectively. Meanwhile, the probe was combined with the hydrogel to construct a visual sensing intelligent test strip, which realized the monitoring of food freshness. In addition, a smartphone device assisted by multiple machine learning methods was used to text Hg2+ and sulfide in real samples. It can be concluded that the fabulous stability, sensitivity, and practicality exhibited by this sensing mechanism give it unlimited potential for assessing the contents of toxic and hazardous substances Hg2+ and sulfide.

Advances in the ratiometric combination of quantum dots for their use in sensing applications

“…Machine learning was used in recent works for generating algorithms for processing highly precise and accurate data for the detection of analytes. In a very recent work by Lu et al , 144 a polychromatic fluorescent-colorimetric paper sensor was designed. The uniqueness of the work is that it is assisted by a smartphone-based sensor, with simultaneous detection of Hg 2+ and sulphide ions.…”

A graphitic C3N4 nanocomposite-based fluorescence platform for label-free analysis of trace mercury ions

“…While traditional Hg 2+ analytical methods, such as atomic absorption/emission spectrometry and inductively coupled plasma mass spectrometry, can achieve highly sensitive detection of low-dose Hg 2+ , they require expensive instrumentation, complex sample preparation, and long analysis time, which limit practical applications. 6,7 In recent years, emerging Hg 2+ detection methods have included colorimetric analysis, 8 electrochemical analysis, 9,10 enzyme analysis, 11 and fluorescence analysis 12,13 among others. Fluorescence sensors, which offer simple operation, rapid response, low analysis cost and high sensitivity, have been increasingly used to detect Hg 2+ .…”

Deciphering Alloy Composition in Superconducting Single-Layer FeSe1–xSx on SrTiO3(001) Substrates by Machine Learning of STM/S Data

“…Alternatively, atomic scale imaging and spectroscopy by scanning tunneling microscopy (STM) and transmission electron microscopy (TEM) have been key to deciphering chemical disorder and electronic heterogeneity in quantum materials. − Regardless, a visual analysis of STM and TEM images can be challenging for multicomponent alloys, particularly beyond the diluted limit due to chemical disorders and electronic inhomogeneities. One way to address this challenge is to use computational methods/statistical analysis (e.g., fast Fourier transform) to analyze the data obtained from imaging experiments to identify patterns and correlations that may not be immediately apparent through visual inspection alone. − A recent approach is to apply machine learning (ML) algorithms to analyze STM/TEM images, where they can be trained to recognize patterns and correlations in the data . Such an approach has been applied to identify hidden electronic orders in STM imaging of quantum matter and to determine surface chemical bonding and adatom interactions. − …”