Chemiresistive metal oxide gas sensors are widely used for environmental monitoring, industrial safety, and emerging healthcare applications, but their selectivity remains strongly dependent on the rational design of surface catalytic sites. Metal loading is one of the most effective strategies for regulating gas adsorption and surface reactions; however, sensing performance is governed not only by the identity of the catalytic metal but also by its particle size and local coordination environment. Here, we develop a combined theoretical and experimental framework to investigate the effects of catalyst metal identity and particle size on SnO₂-based gas sensors. Density functional theory (DFT) calculations were first employed to establish the relationships among metal coordination, electronic structure, gas adsorption, and sensing selectivity. The calculations reveal that atomic-level dispersion of metal substantially modifies the electronic structure of the active sites compared with nanoparticle, resulting in enhanced interaction with CO and pronounced particle-size-dependent sensing behavior. Based on DFT-derived adsorption and electronic descriptors, a machine-learning model was further constructed to screen transition-metal modifiers and predict metal–gas combinations with favorable target selectivity. Guided by these theoretical predictions, atomic dispersed Pd loaded on SnO2 are predicted as ideal material for selective CO sensing and they are experimentally synthesized and evaluated. The experimental sensing characteristics show good agreement with the theoretical predictions, validating the proposed structure–property relationships. This work establishes an integrated design strategy that connects catalyst identity, particle size, electronic structure, and gas-sensing performance, providing a pathway from empirical catalyst selection toward theory-guided development of selective metal oxide gas sensors.
Project is currently funded by: Member Fees