Supplemental SEM-EDS (MLA) and CT data for the publication "CNN-based 3D characterization and liberation analysis of lithium-bearing slag particles using correlative CT and SEM imaging"

Abstract

This dataset provides the underlying X-ray CT data and processed image stacks used for particle-scale characterization and evaluation of the proposed workflow. The study addresses the challenge of accurately characterizing lithium-bearing slag particles to improve recovery of critical raw materials. A correlative workflow combining 3D CT imaging with 2D SEM-based mineral maps is used to train convolutional neural networks for phase-wise and particle-wise segmentation. This enables scalable 3D characterization with minimal manual labeling effort. The models are applied to particle size fractions of 63–100 µm and 100–250 µm. Results show that conventional 2D approaches systematically overestimate mineral liberation, while the presented 3D approach reduces stereological bias and provides more reliable input for process optimization.