Challenges and solutions in the analysis of micro- and nanoplastics down to 500 nm with automated Raman microspectroscopy: suitable filters, accuracy in the detection, identification, and quantification.
Source: PubMed, NCBI / U.S. National Library of Medicine
Microplastics (1 µm-5 mm) and nanoplastics (<1 µm) are emerging contaminants with potential health implications, requiring reliable detection for risk assessment. Raman microspectroscopy offers combined morphological and chemical information for identifying micro- and nanoplastics (MNPs). This study assesses the feasibility of automated Raman microspectroscopy for detecting and quantifying MNPs down to 500 nm, focusing on suitable filter materials and device precision. Among six tested filters, silicon filters with 1 µm pores (for particles ≥1 µm) and aluminum-coated polycarbonate filters with 0.4 µm pores (Al-PC, for particles ≥500 nm) performed best. They provided strong particle contrast, low background interference, and high Hit Quality Index (HQI) for submicron particles (polystyrene 500 nm beads, median HQI ≈ 91%), outperforming other filters affected by roughness and spectral interference. Automation using open-source software TUM-ParticleTyper 2 with Random Window Sampling enabled unbiased detection and quantification of MNPs down to 500 nm. Several limitations were noted: illumination settings influenced detected particle number and size recognition, with particles oversized by 0.5 µm ± 0.26 µm; stage precision (~100 nm) affected spectral quality and particle number (<100 particles per window) was critical for
Abstract
Microplastics (1 µm-5 mm) and nanoplastics (<1 µm) are emerging contaminants with potential health implications, requiring reliable detection for risk assessment. Raman microspectroscopy offers combined morphological and chemical information for identifying micro- and nanoplastics (MNPs). This study assesses the feasibility of automated Raman microspectroscopy for detecting and quantifying MNPs down to 500 nm, focusing on suitable filter materials and device precision. Among six tested filters, silicon filters with 1 µm pores (for particles ≥1 µm) and aluminum-coated polycarbonate filters with 0.4 µm pores (Al-PC, for particles ≥500 nm) performed best. They provided strong particle contrast, low background interference, and high Hit Quality Index (HQI) for submicron particles (polystyrene 500 nm beads, median HQI ≈ 91%), outperforming other filters affected by roughness and spectral interference. Automation using open-source software TUM-ParticleTyper 2 with Random Window Sampling enabled unbiased detection and quantification of MNPs down to 500 nm. Several limitations were noted: illumination settings influenced detected particle number and size recognition, with particles oversized by 0.5 µm ± 0.26 µm; stage precision (~100 nm) affected spectral quality and particle number (<100 particles per window) was critical for achieving >90% correct material identification. Validation with 500 nm polystyrene beads yielded 67 ± 10% recovery relative to theoretical values. Analysis of potable water samples showed predominantly non-plastic particles, with only 0.36 ± 0.13% in the 0.5-10 µm range identified as plastics, 18% of them <1 µm. Further method development, particularly for sample preparation, will be required for broader application to water and food samples. For interest only in particles ≥1 µm, filters with pore size closer 1 µm (e.g., silicon) are recommended. Overall, automated Raman microspectroscopy can quantitatively analyze MNPs down to 500 nm, supporting improved plastic exposure risk assessment.