2, AFCC Automotive Fuel Cell Cooperation Corp., Burnaby, British Columbia, Canada
Modern electrochemical energy storage and power generation devices depend on microstructures. However, each functional layer or component of these devices has a distinctive characteristic length scale, which leads to increasing heterogeneity in its design. For instance, proton exchange membrane fuel cells (PEMFCs) are being developed as alternative energy sources for both residential and automotive applications. The multi-layered membrane electrode assembly (MEA) has length scales that span six orders of magnitude, from several nanometers of the catalyst particle size in the cathode and anode catalyst layer (CL) to hundreds of micrometers of the carbon fibers in the porous gas diffusion layer (GDL). The design, characterization, and optimization of the structures demands both high resolution and device-scale representativeness, often seen as orthogonal requirements that are difficult, if not impossible, to achieve simultaneously.
Rapid development in three-dimensional (3D) microscopy techniques increasingly answers to the resolution challenge at various scales. Combining with artificial intelligence and high performance computing, the massive amount of 3D imaging data at various scales can be integrated. Multi-scale image-based simulation, as a new characterization workflow, offers the potential of a paradigm change in the characterization of electrochemical energy material systems.
Using PEMFCs as a template system, this presentation reports that Transmission Electron Microscope (TEM) tomography (TEM, 0.6nm resolution), Focused Ion Beam - Scanning Electron Microscope tomography (FIB-SEM, 2.5-10nm resolution) and Micro-Computed Tomography (MicroCT, 0.3-10micron resolution) are correlatively employed to image MEA at different scales. Unified structural characterization is achieved via combining 3D imaging data at multiple scales. An up-scaling approach based on TEM, FIB-SEM and MicroCT reconstruction of catalyst layer , micro-porous layer and porous GDL are developed to accurately predict electrical and fluid transport properties, which compares favorably with experimental and literature data. In addition to the benefit of direct visualization of microstructures at various scales, this approach overcomes various difficulties and challenges from physical experiments. The framework is being applied to different electrochemical energy materials including solid oxide fuel cell, lithium battery, magnesium battery, and solar energy materials.