Microplastics quantification in organic-rich samples: the relevance of testing substrate-specific calibration curves
Résumé
Urban agriculture could rely on waste-based substrates, but potential contaminants such as microplastics should be evaluated for safe public use. However, quantifying macroplastics in these substrates is challenging due to their high organic matter content, which is difficult to remove completely, leading to interfering compounds and unreliable results. This study investigated the underexplored effects of organic matter on microplastics quantification employing pyrolysis-GC-MS. Natural organic matter (NOM) removal methods were tested on organic-rich peat-based substrate, reaching up to 46% of reduction with Fenton’s reaction. Then, calibration curves were prepared in two inorganic matrices, silicon dioxide and glass fiber powders, for high density polyethylene (HDPE), polyethylene terephthalate (PET), polypropylene (PP), polystyrene (PS) and polyvinyl chloride (PVC) across nine concentrations, from 0.01 to 10 µg/mg, adding polyflurostyrene (PFS) as internal standard. The selectivity of several polymer pyrolytic markers was compared. Polymer-spiked samples were subjected to Fenton’s oxidation and quantified with both inorganic calibration curves, overestimating polymer contents, up to four times for PET and PVC. The preparation of a third calibration curve, specific for peat, improved results for PS, but not for PP, PET, and PVC. For the first time, the three calibration curves were tested on untreated polymer-spiked (HDPE, PP, PS) waste-based substrates, and resulted in a better estimation closer to the expected polymer concentrations when substrates closely matched the curve’s matrix composition. The comparison of three calibration curves made with different matrices showed that the quantification of plastic polymers in organic-rich samples could be improved using matrix-specific calibration curves even without a complete NOM removal. This represents a novel methodological approach for plastic polymers quantification in complex matrices, minimizing the sample pre-treatment that could cause the loss of nanoparticles during filtration, evidencing that matrix similarity is key for reliable quantification in NOM-rich samples, even without its complete removal.