Machine learning filters out efficient electrocatalysts in the massive ternary alloy space for fuel cells

High-throughput design of bimetallic core–shell catalysts for the electrochemical nitrogen reduction reaction

“…The maximum limit of the effective strain effect in the core–shell catalysts is known to be around 5 atomic layers. 39,40 Similarly, a recent experiment on Cu 3 Au–Pt core–shell catalysts for fuel cell applications 41 shows that the strain effect of the Pt shell layers is maintained up to 1.6 nm. Obviously, the strain effect is maximized on mono-layered shells which can be also readily synthesized in experiments.…”

Towards Reliable Prediction of Performance for Polymer Electrolyte Membrane Fuel Cells via Machine Learning-Integrated Hybrid Numerical Simulations

“…4.2.6. ML in the Field of CL Machine learning approaches have been used in CL for feature extraction [244], optimization [58,147,245,246], predicting performance [247], and degradation of CL on PEMFC [156,[248][249][250][251][252]. Wang et al [244] implemented deep learning super-resolution and multi-label segmentation to process the images from X-ray micro-computed tomography, followed by LBM with multi-relaxation time (MRT) for water management modeling.…”

“…Moreover, the suggested model proves to be effective for optimizing electrocatalyst performance and prediction modeling, with an impressive R 2 of 99.99%. Figure 7e shows a framework for quantitative analysis and accurate prediction, proposed by Yao et al [246], to improve the design efficiency of CL. A combination of the response surface method (RSM) and ANN is utilized to investigate the effect of CL composition on the performance of PEMFC regarding current density, thermal, and water management.…”