Search for: All records

3D imaging of porous materials in polymer electrolyte membrane (PEM)-based devices, coupled with in situ diagnostics and advanced multi-scale modelling approaches, is pivotal to deciphering the interplay of mass transport phenomena, performance, and durability. The characterization of porous electrode media in PEM-based cells encompassing gas diffusion layers and catalyst layers often relies on traditional analytical techniques such as 2D scanning electron microscopy, followed by image processing such as Otsu thresholding and manual annotation. These methods lack the 3D context needed to capture the complex physical properties of porous electrode media, while also struggling to accurately and effectively discriminate porous and solid domains. To achieve an enhanced, automated segmentation of porous structures, we present a 3D deep learning-based approach trained on calibrated 3D micro-CT, focused ion beam-scanning electron microscopy datasets, and data from physical porosity measurements. Our approach includes binary segmentation for porous layers and a multiclass segmentation method to distinguish the microporous layers from the gas diffusion layers. The presented analysis framework integrates functions for pore size distribution, porosity, permeability, and tortuosity simulation analyses from the resulting binary masks and enables quantitative correlation assessments. Segmentations achieved can be interactively visualized on-site in a 3D environment. 

The rapidly growing use of imaging infrastructure in the energy materials domain drives significant data accumulation in terms of their amount and complexity. The applications of routine techniques for image processing in materials research are often ad hoc , indiscriminate, and empirical, which renders the crucial task of obtaining reliable metrics for quantifications obscure. Moreover, these techniques are expensive, slow, and often involve several preprocessing steps. This paper presents a novel deep learning-based approach for the high-throughput analysis of the particle size distributions from transmission electron microscopy (TEM) images of carbon-supported catalysts for polymer electrolyte fuel cells. A dataset of 40 high-resolution TEM images at different magnification levels, from 10 to 100 nm scales, was annotated manually. This dataset was used to train the U-Net model, with the StarDist formulation for the loss function, for the nanoparticle segmentation task. StarDist reached a precision of 86%, recall of 85%, and an F1-score of 85% by training on datasets as small as thirty images. The segmentation maps outperform models reported in the literature for a similar problem, and the results on particle size analyses agree well with manual particle size measurements, albeit at a significantly lower cost.