Machine-Learning Exploration of Transition-Metal-Doped SnS2 Nanomaterials for Sensing C5F10O/CO2 Decomposition Products

Abstract With growing concern over greenhouse gas emissions, C5F10O/CO2 is a promising SF6 alternative, yet its decomposition under partial discharge produces harmful CF4, CO, and COF2, threatening equipment safety. Developing efficient sensing materials for these gases is crucial. In recent years, significant progress has been made in the field of nanomaterials for sensing applications. Here, we propose a high-throughput strategy combining density functional theory with machine learning (ML) focusing on two-dimensional SnS2 nanosheets as the material substrate. Through systematic feature engineering, including cross-validation of the Pearson correlation coefficient, Spearman’s rank correlation coefficient, and the maximum information coefficient, an optimal feature set was constructed. On this basis, an optimal ML model was established to predict the interaction strength between the gas products and transition-metal-doped SnS2. Applying multiple screening criteria identified the most promising sensing materials. Electronic structure analysis further revealed the nanoscale sensing mechanisms. This work provides a highly efficient method for identifying nanoscale gas-sensitive materials and advances the development of environmentally friendly online monitoring technology for inert gases.