Recognition of NO2 and O3 gases using patterned Cu2O nanoparticles on IGZO thin films through machine learning

Abstract

Using different nanoparticles (NPs) in gas sensor arrays is a common method for enhancing gas selectivity. However, gas sensor array systems are highly complex and require large working area. This study explores a simple solution process for fabricating patterned Cu₂O NPs on an amorphous indium gallium zinc oxide (a-IGZO) thin film, aimed at the selective detection of nitrogen dioxide (NO₂) and ozone (O₃) gases. The novel device consists of pure a-IGZO and Cu₂O NPs decorated a-IGZO, which effectively increases the distinctive features of the sensor responses. We employed various machine learning algorithms, including support vector machines (SVM), k-nearest neighbors (KNN), naive Bayes (NB), random forest (RF), and linear discriminant analysis (LDA), to analyze the sensor responses, achieving high prediction accuracy. This method can be adapted for the fabrication of other metal oxide semiconductor-based sensors, potentially broadening the scope of applications in gas sensing and environmental monitoring.