This work achieved the chemical discrimination of benzene series (toluene, xylene isomers, and ethylbenzene gases) based on the Ti-doped Co3O4 sensor. Benzene series gases presented different gas-response features due to the differences in redox rate on the surface of the Ti-doped Co3O4 sensor, which created an opportunity to discriminate benzene series via the algorithm analysis. Excellent groupings were obtained via the principal component analysis. High prediction accuracies were acquired via k-nearest neighbors, linear discrimination analysis (LDA), and support vector machine classifiers. With the confusion matrix for the data set using the LDA classifier, the benzene series have been well classified with 100% accuracy. Furthermore, in situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and density functional theory calculations were conducted to investigate the molecular gas-solid interfacial sensing mechanism. Ti-doped Co3O4 showed strong Lewis acid sites and adsorption capability toward reaction species, which benefited the toluene gas-sensing reaction and resulted in the highly boosted gas-sensing performance. Our research proposed a facile distinction methodology to recognize similar gases and provided new insights into the recognition of gas-solid interfacial sensing mechanisms.
Keywords: Ti-doped Co3O4; algorithm analysis; in situ DRIFTS; sensing mechanism; xylene isomers.