DFT-assisted multiscale descriptor analysis of gas adsorption behavior on rare-earth-doped SnO

Abstract

Gas sensors play a critical role in environmental monitoring, industrial safety, and medical diagnostics, highlighting the need for efficient and cost-effective sensing materials. By combining machine learning (ML) with density functional theory (DFT), we systematically investigated the adsorption mechanisms of target gases (CH4, H2, CO, and NH3) on rare-earth-doped SnO2 surfaces (Ln-SnO2). Our work successfully predicted the adsorption energies of target gases across various doped systems and identified the important electronic and structural factors influencing adsorption strength based on a random forest model. Furthermore, equations derived from the sure independence screening and sparsifying operator (SISSO) provided clear and interpretable structure-energy relationships. The results indicate that adsorption strength is jointly determined by molecular charge transfer and geometric relaxation under the constraint of the surface work function (Wf). Through analyses of recovery time and sensitivity, we clarified the response behaviors of rare-earth dopants toward different target gases. These findings deepen our understanding of Ln–SnO2–gas interactions.