AI-Agent-Guided Design of Dual-Scale Modulated Nickel-Based Catalyst with Built-In Electric Field for Enhanced Biomass Electrooxidation

Electrochemical synthesis offers a sustainable route for chemical production, and integrating artificial intelligence (AI) into electrocatalyst design promises to accelerate the development of efficient processes. Herein, we report an AI-agent-assisted strategy for the design of a nickel-based catalyst with a built-in electric field (BEF) for the electrooxidation of 5-hydroxymethylfurfural (HMF) to 2,5-furandicarboxylic acid (FDCA). The AI-agent autonomously identified Mn doping as a means to create a BEF that simultaneously optimizes the electronic structure and the interfacial microenvironment. These synergistic effects enable the resulting Mn–Ni(OH) 2 catalyst to achieve a current density exceeding 700 mA cm –2 at 1.45 V vs RHE, with HMF conversion, Faradaic efficiency, and FDCA selectivity all surpassing 99%. Notably, after 45 cycles, the activity of Mn–Ni(OH) 2 remains stable. When assembled into a flow electrolyzer, current densities of 0.5 and 1 A cm –2 are achieved at cell voltages of 1.735 and 2.162 V, respectively, over a duration of 100 h. Characterization and simulation results reveal that the BEF enhances charge transfer by modulating the Ni e g * orbital and accelerates mass transport by disrupting the interfacial hydrogen-bond networks. This AI-assisted, dual-regulation strategy bridges the gap between catalyst electronic structure engineering and interfacial microenvironment design for sustainable electrosynthesis applications.