Characterizing the Influence of Filter Substrates on Raman Measurements of Microplastics

Raman spectroscopy is a powerful tool for the characterization of microplastics in environmental samples; however, measurements are typically performed directly on filter substrates, making substrate selection a critical factor for reliable identification. Here, we systematically investigate the influence of 18 commonly used filter substrates on Raman-based microplastic analysis using polystyrene particles with diameters of 7 and 1 μm as model systems. By combining confocal Raman microscopy, spatial Raman mapping, substrate background correction, and PCA-SVM-based classification, we demonstrate that substrate suitability is governed primarily by the geometric relationship between particle size and the confocal sampling volume. For particles substantially larger than the confocal point spread function, Raman spectra are sample-dominated and can be reliably acquired on a wide range of substrates, including polymeric membranes. At particle sizes approaching the confocal sampling volume, substrate contributions become unavoidable; however, mapping-based localization and local background subtraction enable reliable recovery of microplastic spectra even on chemically interfering polymeric filters. These findings challenge the common assumption that only nominally Raman-silent substrates are suitable and provide a practical, physically grounded framework for cost-efficient and flexible filter selection in Raman-based microplastic analysis. Furthermore, this work provides a practical substrate-selection guide for confocal Raman microspectroscopy across different sample types measured on supporting substrates.