High-Throughput Microplastic Sizing and Quantification in Water Using Static Light Scattering and Machine Learning

Quantitative analysis of microplastic (MP) size and concentration is essential for a better understanding of plastic pollution in aquatic environments. Conventional single-particle methods, however, are often time-intensive and can be cumbersome for high-throughput analysis. In this study, we present a novel measurement approach combining Static Light Scattering (SLS) with machine learning (ML) algorithms to enable a rapid and accurate analysis of the size and concentration of microplastics in water samples. Polystyrene microplastics ranging from 0.5 to 20 mu m are used to evaluate the technique. We employ Principal Component Analysis (PCA) to reduce the dimensionality of scattering intensity patterns, facilitating feature extraction for machine learning classification. We automate the classification process with machine learning approaches such as k-Nearest Neighbors (KNN), Multilayer Perceptron (MLP), Support Vector Machines (SVM), Bernoulli Naive Bayes (BNB), Decision Trees (DT), Gradient Boosting (GB), Linear Discriminant Analysis (LDA), Linear Regression (LR), and Random Forest (RF). These models are fine-tuned and optimized to accurately determine the size distribution of MPs in water, with KNN and MLP providing more than 96% classification accuracy, outperforming the others. Concentration is subsequently quantified through a linear fitting model based on the specified particle sizes. Our integrated SLS-ML approach offers a scalable, high-throughput alternative to traditional methods after filtering/preprocessing the water, enhancing the efficiency of microplastic monitoring. This technique provides a valuable tool for advancing environmental assessments and offers potential for broader applications in pollution tracking in water and water quality management.