Enhanced hydrogen adsorption on boron nickel gold modified Si60 nanocluster via DFT and machine learning analysis

In this study, we investigated the catalytic potential of silicon-based nanoclusters, specifically Si₆₀, for hydrogen evolution reactions (HER), focusing on enhancements via substitutional doping with nickel (Ni) and boron (B), where boron and nickel are incorporated into the Si₅₉ lattice, enabling electronic structure modulation. Using Density Functional Theory (DFT) with the B3LYP functional, we examined hydrogen adsorption behavior on boron- doped, Ni-doped, gold-encapsulated Si₅₉ nanoclusters, denoted as B_x^dopNi^dopAu^encSi₅₉ (x = 1, 2, 3). Our findings show that doping introduces minimal structural distortion while improving cluster stability and reactivity. Boron-doping notably reduces the energy gap, enhancing electron transfer and promoting hydrogen adsorption. Gibbs free energy analyses confirm somewhat favorable catalytic activity, with ΔG_H values ranging from - 0.837 to - 0.848 eV, highlighting the B_x^dopNi^dopAu^encSi₅₉ engineered doped nanoclusters as promising catalysts. Additionally, machine learning models, particularly ElasticNet Regression (R² = 0.9942), accurately predict hydrogen adsorption energies across various surfaces, from pristine H₂@Si₆₀ to B₃- doped systems. This demonstrates the models' capability to capture structure-property relationships, accelerating catalyst optimization. Computational screening of Si₅₉-based nanostructures revealed that H₂@B₂^dopNi^dopAu^encSi₅₉ exhibits a ΔG_H value of - 0.836 eV, suggesting comparatively balanced hydrogen adsorption and improved HER catalytic potential among the evaluated systems. Overall, the results suggest that precise doping and surface engineering can significantly enhance the electronic, storage, and catalytic properties of silicon nanoclusters, offering valuable insights for the design of efficient HER catalysts in future sustainable energy technologies.