We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Optimizing spectral classification and oxidation estimation of environmental Microplastics
Summary
Researchers performed an intercalibration exercise to optimize spectral classification and oxidation estimation methods for weathered environmental microplastics, finding that spectral libraries built from weathered particles improve polymer type identification accuracy compared to libraries based on pristine reference materials.
Microplastics (MPs) research has been challenging due to the wide range of methodologies that have been applied for sampling, extraction, counting and polymer type identification. Identifying MPs with microscopy only can yield to false positive and false negative results. Thus, polymer type identification is required to confirm that particles classified as MPs are indeed plastic polymers. The accurate identification of the small and weathered microplastics found at sea is critical. Therefore building spectral libraries based on weathered MPs is important. Here, we performed an intercalibration exercise using 200 spectra of sea surface floating microplastics (¿300μm) applying two different FTIR spectrometers. Differences and similarities of the results were investigated using various spectral preprocessing methods and tools and checked against various libraries. In addition, for a total of 2000 spectra from marine floating microplastics the carbonyl index (CI) of polyethylene (PE) polypropylene (PP) and Polystyrene (PS), including a diversity of sizes, and shape types, was calculated, using two different CI measurement techniques as described in the literature. The CIs of marine floating MPs were compared with those from literature data from accelerated ageing experiments and from MPs of known 'age'. CI calculation of the specified area under band (SAUB), revealed high variability ranging from 0.04 to 5.98 for the PE and 0.06 to 3.96 for the PP. The percentage of PE particles with SAUB ratio between 0.04 - 0.44 was 28 Also see: https://micro2024.sciencesconf.org/559692/document
Sign in to start a discussion.
More Papers Like This
Optimizing spectral classification and oxidation estimation of environmental Microplastics
Researchers conducted an intercalibration exercise using 200 FTIR spectra from sea surface floating microplastics analyzed on two different spectrometers, comparing spectral preprocessing methods and library-matching tools to assess identification reliability for weathered environmental particles. They also calculated the carbonyl index for 2,000 spectra from marine floating microplastics across multiple polymer types, finding high variability in oxidation levels that complicates comparisons with accelerated aging experiments.
Enhanced Identification of Weathered Plastics Through the Improvement of Infrared Spectral Libraries
Researchers developed an improved infrared spectral library specifically designed to identify weathered and degraded plastics that conventional libraries often misidentify. The new library increased match rates by 7.3% for thermally oxidized plastics and improved identification of mechanically abraded samples, addressing a significant gap in accurate microplastic detection and environmental risk assessment.
Machine learning based workflow for (micro)plastic spectral reconstruction and classification
A machine learning pipeline combining two spectral reconstruction models with four classification algorithms can identify microplastic polymer types from spectral data with up to 98% accuracy on processed spectra. Applied to real environmental samples, the best model achieved 71% top-one accuracy and over 90% top-three accuracy. Automated, high-accuracy microplastic identification tools are critical for scaling up environmental monitoring and making large-scale surveys practical.
Impact of weathering on the chemical identification of microplastics from usual packaging polymers in the marine environment
The impact of environmental weathering on the chemical identification of common microplastics was investigated, examining how UV radiation, mechanical abrasion, and microbial activity alter the spectroscopic signatures used for polymer identification. Weathered plastics were harder to correctly identify than pristine ones, highlighting the need for reference libraries that include aged material.
Machine Learning of polymer types from the spectral signature of Raman spectroscopy microplastics data
Machine learning models were applied to Raman spectroscopy data to improve polymer type identification in environmentally weathered microplastics, which are harder to classify than pristine samples. The approach achieved better accuracy by accounting for spectral changes caused by UV exposure and physical degradation.