Machine learning-assisted electronegativity-driven electronic modulation of high-entropy Prussian blue analogues for boosting performances of Li–CO2 batteries

High-entropy materials have emerged as promising catalysts for next-generation energy storage systems owing to their unique structural stability and tunable electronic environments. However, vast compositional space and complex element interactions hinder rational design, making experimental trial-and-error both time-consuming and low working efficiency, thereby highlighting the need for predictive strategies to accelerate the discovery of optimal compositions. In this study, a machine learning model is developed to accurately predict the charging voltages of Li–CO 2 batteries with HE-PBA cathodes (MSE = 0.0131 V). From 6189 candidates, PBA-FeZnInNiCu is identified as a low charging voltage. Feature analysis reveals a strong electronegativity-voltage correlation, which is experimentally validated by synthesizing and systematically comparing PBA-FeZnInNiCu with PBA-FeMnInNiCu and PBA-FeZnCoNiCu. Benefiting from synergistic multi-metal interactions and charge redistribution, PBA-FeZnInNiCu delivers ultrahigh discharge capacity, stable cycling and a low charging plateau. Electronegativity-driven electron redistribution combined with high-entropy-induced electronic coupling optimizes the active-site electronic configuration, enhances adsorption/desorption of intermediates, and accelerates CO 2 reduction/evolution kinetics. This work highlights the key role of electronegativity regulation and composition tuning in boosting the performance of high-entropy cathode catalysts for Li–CO 2 batteries. • Machine-learning-assisted screening identifies optimal compositions of high-entropy Prussian blue analogue (HE-PBA) cathodes for Li–CO 2 batteries. • Machine learning reveals electronegativity as a key descriptor for designing efficient HE-PBA cathodes. • Electronegativity differences induce electron redistribution and modulate the electronic structure of catalytic active sites. • The optimized HE-PBA catalyst (PBA-FeZnInNiCu) enables low charging voltage and improved CO 2 redox kinetics.