Comparing Automated Methods for Analyzing Large Data Sets in Microplastics Research: A First-of-its-Kind Study

For the first time, researchers at the University of Bayreuth and the Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research (AWI) have compared two automated analysis methods for microplastics data to determine the uniformity of the analytical standards used in the field. The study, published in the journal Analytical and Bioanalytical Chemistry, showed significant deviations between the two methods, particularly for small particles with a high potential for hazards. The researchers suggest that analytical standardisation is necessary to ensure comparability and reliable data in the field of microplastics research.

Microplastics refer to plastic particles that are less than 5 millimetres in size. In the early days of research, microplastics were identified using visual criteria alone, which could lead to erroneous results. Today, the identification of microplastics is carried out using chemical analysis. Micro-Fourier transform infrared spectroscopy, or micro-FTIR spectroscopy, is currently one of the most reliable methods used to measure microplastics.

To analyse microplastic particles that are smaller than 0.5 millimetres, samples are applied to filters and analysed using micro-FTIR spectroscopy. However, the measurement produces up to several million FTIR spectra, which makes manual analysis for microplastics impossible. Therefore, reliable automated computer methods are necessary for the analysis.

For automatic analysis of FTIR data sets, various evaluation algorithms are used. In the study, two evaluation algorithms were compared: the systematic identification of MicroPlastics in the environment (siMPle) analysis tool, and the Bayreuth Particle Finder (BPF). Both algorithms have the advantage of examining the entire sample, avoiding biases that arise when only parts of a sample are selected.

The researchers compared two sets of samples using both evaluation algorithms. The quantity and size of the microplastic particles and the proportions of various polymers were measured. One sample set contained ten water samples from the Upper and Middle Weser, while the other set contained ten water samples from the Lower/Outer Weser and the Jade Bay.

While the results were largely consistent, deviations were found in the range of particles smaller than 50 micrometres, where the algorithms could make wrong decisions due to poorer FTIR spectra quality. The researchers suggested that further comparative research is needed to identify microplastic particles of all sizes without error using automated methods, and to ensure that the data obtained using these methods are subjected to critical review.