Multivariate spectral analysis in quartz-enhanced photoacoustic spectroscopy

Here we report on the application of multivariate analysis on optical sensors for gas detection based on Quartz-Enhanced Photoacoustic Spectroscopy (QEPAS) technique, focused on the analysis of complex gas mixtures. In real-world applications the effects of spectral and non-spectral interference occurring within the gas samples cannot be neglected in order to increase sensors selectivity and accuracy. In this work, Partial Least Squares Regression (PLSR) is selected as regression technique and tested on different gas samples for different applications. PLSR is able to retrieve analytes concentrations filtering out both: i) spectral contributions of analytes characterized by strongly overlapping features; ii) correlation effects due to the interaction among the sample’s components, i.e., matrix effects characterizing the photoacoustic detection.