Technological innovations in the field of materials science are often connected to the nanometric scale material evaluation which requires utilizing a combination of advanced characterization techniques. The extensive data acquired from these characterization techniques necessitates meticulous and efficient analysis, which can only be accomplished by automated digital image and data processing techniques. The development of such techniques will aid further advancements in the design of functional materials. In this work, we report a novel automatic analysis approach using data acquired by the Scanning Transmission Electron Microscopy (STEM) and Energy Dispersive Spectroscopy (EDS) to determine the component distribution and microstructure descriptors for an anion exchange membrane fuel cell (AEMFC). The current research aims to validate the automated approach by comparison to a manual image processing method based on ImageJ reported in a previous study [1]. STEM and EDS data for three different Pd/CeOx-based catalyst samples with varying bulk atomic fractions of Ce/Pd were collected and analyzed. The STEM/EDS data was used to investigate how the interfacial contact area and agglomerate size distributions of Pd and Ce can be related to the microstructural-level interactions and processes and potentially affect the electrochemical performance of an anion exchange membrane fuel cell (AEMFC). The comparison highlighted the strengths and limitations of both automated and manual approaches and offered substantial justification to predict how excessive Ce agglomeration can be linked to a reduction in the electrochemical performance of the fuel cell.

[1] R. K. Singh et al., “Synthesis of CeOx-Decorated Pd/C Catalysts by Controlled Surface Reactions for Hydrogen Oxidation in Anion Exchange Membrane Fuel Cells,” Adv. Funct. Mater., vol. 30, no. 38, 2020, DOI: 10.1002/adfm.202002087.