Abstract

In research laboratories, data generated by instruments is constantly increasing. Their analysis can become extremely complex without relevant mathematical tools and computational power. In physical chemistry, a discipline called Chemometrics has been developed to acquire a better knowledge of the samples analyzed. It brings together data reduction, classification methods, regression methods and signal unmixing. Nowadays, the interest in nanoscience is important in many research fields and the instrumentation must always be more efficient to answer future challenges. Hyperspectral imaging techniques are valuable tool to analyze complex samples and provide significant molecular information. The coupling spectrometers with microscopes make possible maps generation that represent the spatial distribution of chemical components from a sample. Nevertheless, the diffraction limit dictated by the photon wavelength becomes a real constraint when submicron-sized samples are analyzed. In this chapter, we present an original method which use both super-resolution and multivariate curve resolution in confocal Raman imaging to break the instrumental limits in order to characterize atmospheric aerosols.