Search

Abstract Recent advances in 3D printing have enabled the manufacture of porous electrodes which cannot be machined using traditional methods. With micron-scale precision, the pore structure of an electrode can now be designed for optimal energy efficiency, and a 3D printed electrode is not limited to a single uniform porosity. As these electrodes scale in size, however, the total number of possible pore designs can be intractable; choosing an appropriate pore distribution manually can be a complex task. To address this challenge, we adopt an inverse design approach. Using physics-based models, the electrode structure is optimized to minimize power losses in a flow reactor. The computer-generated structure is then printed and benchmarked against homogeneous porosity electrodes. We show how an optimized electrode decreases the power requirements by 16% compared to the best-case homogeneous porosity. Future work could apply this approach to flow batteries, electrolyzers, and fuel cells to accelerate their design and implementation.

Abstract Lattices remain an attractive class of structures due to their design versatility; however, rapidly designing lattice structures with tailored or optimal mechanical properties remains a significant challenge. With each added design variable, the design space quickly becomes intractable. To address this challenge, research efforts have sought to combine computational approaches with machine learning (ML)-based approaches to reduce the computational cost of the design process and accelerate mechanical design. While these efforts have made substantial progress, significant challenges remain in (1) building and interpreting the ML-based surrogate models and (2) iteratively and efficiently curating training datasets for optimization tasks. Here, we address the first challenge by combining ML-based surrogate modeling and Shapley additive explanation (SHAP) analysis to interpret the impact of each design variable. We find that our ML-based surrogate models achieve excellent prediction capabilities (R 2 > 0.95) and SHAP values aid in uncovering design variables influencing performance. We address the second challenge by utilizing active learning-based methods, such as Bayesian optimization, to explore the design space and report a 5 × reduction in simulations relative to grid-based search. Collectively, these results underscore the value of building intelligent design systems that leverage ML-based methods for uncovering key design variables and accelerating design.

Abstract Multiphysics modeling enables probing of conditions inside a CO2 electrolyzer that are difficult to measure, such as local concentrations and pH, as well as rapid testing of possible design changes. A one‐dimensional model for a zero‐gap membrane electrode assembly (MEA) CO2 electrolyzer was developed with the assumption that catalyst layers interact with the membrane ionomer such that the ionomer affects the underlying kinetics. The kinetics for bicarbonate reacting to form hydrogen are fit using a planar electrode model for silver with an ionomer coating. The MEA model results are validated against experimental studies for current density and product selectivity. Flooding of the cathode is modeled using saturation curves, and results show that blocked pores in the microporous layer play a significant role in limiting the mass transport at high potentials (>2.8 V). Sensitivity studies showed that CO Faradaic efficiency can be increased by decreasing catalyst layer thickness and porosity, and decreasing KHCO3 concentration.

Abstract Invited for this issue's Front Cover are researchers from the Carbon Initiative at Lawrence Livermore National Laboratory and the SUNCAT Center at Stanford University. The front cover shows a cross‐section of the cathode of a membrane electrode assembly for CO2 electrolysis looking down the feed channel. CO2 molecules flow down the channel and diffuse up through the gas diffusion layer to the silver catalyst, where CO2 reacts to form CO, while also competing against hydrogen reduction from water. Some of the CO2 molecules react to form bicarbonate and carbonate ions, which can diffuse across the membrane, where they react at the anode to form CO2 again. The top of the image shows CO2 that has crossed over through the membrane into the anolyte. Cover design by Brendan Thompson. Read the full text of the Research Article at 10.1002/celc.202300566.