Impact of Soil Matrix on Microplastics Analysis by Py-GC/MS

Analyzing Microplastics in the Environment
Poster Presentation

Prepared by Y. Hiramatsu
Shimadzu Scientific Instruments, 7102 Riverwood Dr, Columbia, Maryland, 21046, United States


Contact Information: [email protected]; 240-749-3572

ABSTRACT

Microplastic (MPs) analysis is becoming more prevalent due to the environmental impact these pollutants have on the environment and human health. Recent studies using Py-GCMS for the analysis of MPs in human tissues and other complex matrices have published controversial results due to the overlapping markers of the pyrolizates from polymers and matrix components. Hence, this complicates identification and quantification of MPs and increases the risk of reporting false positives. In this work, we validated the workflow for the unequivocal identification of MPs in soils. Three soil samples from different locations were homogenized, dried at 40 °C for 3 hours, and analyzed directly (2 mg) by Py-GC/MS. A 5-point calibration was generated in triplicate using the Frontiers Low Mix Standard for 10 polymers: PE, PP, PS, ABS, SBR, PMMA, PC, PVC, N6, and N-66. Calibration curves showed strong linearity for all ten polymers (R² > 0.9927), despite differences in sensitivity. Short-term repeatability at the lowest calibration level yielded acceptable %RSD values (<28%). Polymers were quantified on the three samples using external calibration curves generated from known standards. Background signal intensity varied among samples, with one soil exhibiting substantially elevated background. Polyvinyl chloride (PVC) was also detected in the blank samples with >90% spectral match, indicating matrix-derived interference. Consequently, recovery studies were conducted using blank samples spiked with 2 mg of a polymer standard mixture (CaCO₃ diluent) and analyzed in triplicates. Recovery percentages ranged from 50 to 102% for most polymers, except for PVC that was between 150 – 350%xxx. Polymer identification requires targeting one or more unique pyrolyzates from a spectral library. Further investigation of the samples revealed that specific PVC pyrolyzates were not identified. Hence, PVC identifications were false negatives, but the measured ions did interfere with how the PVC is being identified by the software. These results demonstrate that matrix composition significantly impacts polymer detection, quantification, and confidence in spectral assignments. In this study, a matrix calibration curve was applied to the samples to counteract interferences and increase the confidence of the results. To further optimize workflows for MPs analysis, environmentally realistic reference materials are critically needed, including spiked soils with multiple polymers and aged or weathered MPs rather than pristine lab standards.