上海科技大学知识管理系统

This study focuses on the development of a sustainable and highly sensitive Antheraea pernyi (A. pernyi) silk paper-based electrochemical sensor, utilizing the excellent solvent resistance of silk paper as the substrate and incorporating molybdenum disulfide (MoS2) functional units. The electrochemical sensing principle of this sensor is based on a transpiration-driven electricity generation mechanism. Different solvents interact distinctly with the sensor, resulting in transpiration-driven electrical signals with varying characteristics. Although these electrical signals display complex band-shaped patterns lacking sharp peaks or specific mutation points, they exhibit highly reproducible trends for specific solvent systems. Consequently, this study further employs a feedforward neural network to construct a solvent identification model based on the oak silk fibroin ultrafine paper electrochemical sensor. This model processes and analyzes these complex yet highly repetitive electrical signals. The AI-driven A. pernyi silk paper-based electrochemical sensor not only accurately identifies solvents with highly similar chemical structures and properties, such as methanol, ethanol, isopropanol, and deionized water, but also recognizes chemically similar mixed solvent systems, like varying proportions of ethanol-water mixtures. In 100 identification tests, the accuracy was 100%. The development of this AI-driven A. pernyi silk paper-based electrochemical sensor offers a convenient, cost-effective method for rapid and accurate solvent detection, holding significant implications in fields such as alcohol testing, environmental monitoring, and chemical analysis.