Artificial intelligence-assisted design of FeNi-LDH-NiTe needle-like array for high current density oxygen evolution reaction

Active, stable, and low-cost alkaline oxygen evolution reaction (OER) catalyst is essential for large-scale hydrogen (H 2 ) production via electrochemical water splitting. Traditional catalyst design often relies on experience-driven, trial-and-error approaches, limits exploration efficiency. To overcome this, we present an artificial intelligence (AI)-assisted high-throughput framework combining large language models (LLMs)-based literature mining and Python-driven simulations. LLMs identifies FeNi-layered double hydroxide (FeNi-LDH) has excellent activity for OER while Python-based simulations propose a needle-like facilitate bubble detachment. In our experimental, the as-prepared FeNi-LDH exhibits good OER activity (237 mV at 10 mA cm −2 ). After growing FeNi-LDH on NiTe array, the needle-like FeNi-LDH-NiTe not only delivers enhanced activity (198 mV at 10 mA cm −2 ), but also exhibits faster bubble generation and migration, which allow it to stably work at high current density (1 A cm −2 ). The successful AI-guided catalyst design proposes a clear strategy to decouple kinetic and non-kinetic optimization for exploring excellent electrocatalyst.