polymer differentiation and dynamic monitoring Ghazal, Naina creator Xue, Wenchao Chairperson Cruz, Simon Guerrero Examination Committee Chantri Polprasert Examination Committee PMU-KPCIP-AIT Scholarship Scholarship Donor text series technical report th Pathum Thani, Thailand Asian Institute of Technology 2026 continuing eng

print

85 leaves : ill.+ 1 online resource The Electrical Sensing Zone (ESZ) method is an emerging approach for microplastic detection, capable of providing particle count and size information. This study extends its application beyond conventional counting and sizing by integrating ESZ with machine learning to enable polymer differentiation and real-time monitoring of particle interactions.For polymer differentiation, ESZ signal features were used to classify three polymer types, polyethylene (PE), polyamide (PA), and polyvinyl chloride (PVC), based on particle transit dynamics. Pulse morphology study showed a broader pulse for PE compared to PA and PVC inferring that buoyancy induced pulse broadening provides an excellent discrimination between PE and PA/PVC, however, density induced difference in the pulse width between PA and PVC was indiscernible based on morphology analysis alone. A Random Forest classifier achieved an overall accuracy of 91.24% in three-class discrimination, with high precision for PE and moderate misclassification between PA and PVC due to similar density characteristics.ESZ was applied for real-time monitoring of aggregation dynamics between PVC microplastics and powdered activated carbon (PAC) over a four-hour experiment comprising six phases and 24,557 particle events. While the median particle diameter remained approximately constant at 111.7 æm, the distribution width (Span) increased from 0.265 to 0.415 during early flocculation, indicating aggregate formation. A three-component Gaussian Mixture Model (GMM) successfully differentiated PAC homoaggregates, PVC particles, and PAC-PVC heteroaggregates. The heteroaggregate fraction peaked at 23% during early flocculation before declining due to settling, indicating effective removal.The results demonstrate that ESZ, when coupled with machine learning, can move beyond conventional particle sizing to provide simultaneous insight into microplastic identity and dynamic behavior, offering strong potential for scalable, in-situ environmental monitoring and water treatment applications. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Environmental Engineering and Management Thesis (M. Sc.) - Asian Institute of Technology, 2026 Coulter principle Microplastics Analysis Microplastics Environmental aspects Environmental monitoring Asian Institute of Technology. http://203.159.5.9/ait-thesis/detail.php?q=B24448 http://203.159.5.9/ait-thesis/detail.php?q=B24448 260616 20260818113150.0