Microplastic Identification Using Impedance Spectroscopy and Machine Learning Algorithms

Abstract

Detecting and classifying microparticles in water and other liquid substances is crucial due to their detrimental impact on ecosystems and human health. This is because particles such as microplastics, micropollutants, or heavy metals in water have demonstrated a high impact on the health of ecosystems and a high risk when this water is used for human consumption. Water quality is a critical factor when it comes to human consumption. Currently, some of these pollutants are not correctly detected during water treatment processes or directly in ecosystems, which can carry health risks for humans and animals. From this point of view, the development of tools for detecting these particles is still needed, and research for new strategies for detecting and classifying these microparticles with in situ methods is required. As a contribution to the solution of this problem, this work presents a microplastic detection and classification methodology that uses an electronic tongue system, impedance spectroscopy, and machine learning algorithms for detecting and classifying microplastics. Validation is performed using various sizes of PET (polyethylene terephthalate) microparticles in water to validate the possibility of classification. Results show the advantages of using the methodology, obtaining high accuracy in the classification process.