Explainable ensemble learning reveals site-driven d z 2 orbital occupancy tuning for enhanced hydrogen evolution on metal-doped Ni-loaded BN catalysts

Key Points

• This research aims to optimize hydrogen evolution using ensemble learning and first-principles calculations.
• Integration of first-principles calculations with ensemble learning
• Development of theoretical models and descriptors
• Focus on bimetallic Ni–TM2@BN catalysts
• Enhanced hydrogen evolution activity predicted for metal-doped catalysts
• Identification of optimal orbital occupancy tuning
• Demonstrated efficiency in accelerating catalyst discovery