Influence of Oxygen Vacancy Distribution on CO₂ Hydrogenation: A Case Study of ZnO and In₂O₃

Oxygen vacancies (OVs) on metal oxide surfaces are widely recognized as catalytically active sites; however, the impact of their distribution on the catalytic performance remains underexplored. In this study, we used density functional theory (DFT) calculations combined with a machine learning potential to investigate the distribution of OVs on the ZnO-(10 $\stackrel{-}{1}$ 0) surface and their role in CO₂ hydrogenation. We efficiently analyzed over 700,000 potential OV configurations by reducing them to unique, irreducible structures using the self-developed DefectMaker program. Our results revealed that higher OV concentrations led to the formation of linear OV structures, which, despite their energetic stability, exhibited lower CO₂ hydrogenation efficiency compared to isolated OVs, due to the reduced surface polarization with linear OVs. Additionally, a comparative investigation on In₂O₃ surfaces revealed a scattered distribution of OVs, maintaining the material's catalytic activity in CO₂ hydrogenation. This work provides a deeper understanding of defect engineering in metal oxides for a more efficient CO₂ conversion.