ARCADE · Accelerated Redox Catalysis through AI-driven Design of Electrodes

The way in which new materials for clean energy are designed is changing rapidly. Thanks to powerful computer simulations, scientists can study chemical processes that were previously too complex to examine in detail. Electrocatalysis is one of them, as it plays a central role in producing clean fuels like hydrogen. Today, advances in machine learning (ML) are transforming catalysis studies by bridging the gap between two fields: high-precision quantum calculations, which describe matter at the atomic level, and large-scale classical simulations, which reveal the statistical properties necessary for understanding reactivity. The ARCADE project will leverage these advances to create the first ML framework designed specifically to understand and predict electron transfer reactions, key chemical processes that determine how efficient a material is at driving electrocatalysis. This tool will enable the prediction of the most promising materials for designing improved electrodes, which are essential components of electrochemical reactions driven by sustainable energy. This knowledge will inform the development of more efficient and sustainable hydrogen energy production, tackling a major priority of the European Union in its urgent transition to greener energy sources. ARCADE will address three major scientific challenges to achieve this: (1) incorporating the effect of the electric field created by an electrode into an ML-based simulation framework; (2) developing a reliable, user-friendly method to realistically model electron transfer events; and (3) ensuring the models can be applied to a wide range of material compositions. ARCADE will be hosted at the COSMO laboratory at EPFL under the guidance of Prof. Michele Ceriotti, providing an ideal environment for scientific innovation, collaboration and personal growth. By combining cutting-edge computational methods with urgent societal needs, ARCADE will help to create cleaner energy solutions.