Chemometrics tools for advanced spectroscopic analyses

Any system is described by several variables, often in the form of hidden information, able to describe and explain functional mechanisms for the majority of the processes which can be evaluated analytically only when we consider entire complex datasets. The relationship between those variables is the key to identify and quantify correlations among the parameters describing the data in a strictly model-free manner. In chemometrics one uses mathematical and statistical methods to improve the understanding of chemical information through the correlation of physical parameters or properties to analytical instrument data. This approach is currently used across chemistry, materials science, biology, with a growing impact is the field of spectroscopy. This paper presents the ability of chemometric technique applied to Advanced Spectroscopic Analyses, examples include spectroscopic data collected from both the High- resolution neutron Spectrometer TOSCA, operating at the ISIS pulsed Neutron and Muon Source (UK) and X-ray fluorescence (XRF) spectroscopy. This work demonstrates the high-resolution of the Principal Component Analysis (PCA) to a spectroscopic data-set dealing with the determination of marker bands from Inelastic Neutron Scattering (INS) spectra of a large data- set, the presence of a probably additional transition phase of one globular molecule and evidencing the metallic nature of the black/brownish inscriptions on daily-use textiles used in ancient Egypt. This study will pave the way for the analysis of multi-parametric, high-throughput INS data, now within reach using state-of-the-art chemical neutron spectrometers such as VESPA.