Predicting the 3D microstructure of SOFC anodes from 2D SEM images using stochastic microstructure modeling and CNNs

Abstract The 3D microstructure of solid oxide fuel cell anodes significantly influences their electrochemical performance, but conventional methods for acquiring high-resolution microstructural 3D data, such as focused ion beam scanning electron microscopy, are costly in both time and resources. In contrast, obtaining 2D images, such as from scanning electron microscopy (SEM), is more accessible, though typically providing insufficient information to accurately characterize the 3D microstructure. To address this challenge, we propose a novel approach that predicts the 3D microstructure from 2D SEM images. The presented method utilizes a low-parametric stochastic geometry model to generate virtual 3D microstructures and employs a physics-based SEM simulation tool to obtain the corresponding 2D SEM images. By systematically varying the model parameters, a large dataset can be generated to train convolutional neural networks. By doing so, we can statistically reconstruct the 3D microstructure from 2D SEM images by drawing realizations from the stochastic 3D model using the predicted model parameters. This workflow is quantitatively validated by an error analysis on geometrical descriptors, which shows that 3D microstructures can be predicted with reasonably high accuracy from 2D SEM images. Thus, this approach is a valuable computational tool that additionally circumvents the typically time-consuming segmentation of image data.