Untargeted/targeted methods for quality assessment in hazelnuts, cocoa, and finished products of the confectionery industry

This Ph.D. thesis explores the development and application of targeted and untargeted analytical methods for quality assessment in hazelnuts, cocoa, and finished confectionery products. Using chromatographic techniques such as UHPLC-HRMS and GC-IMS, the research focuses on the analysis of defects in hazelnuts, the discrimination of cocoa origins, and the monitoring of phenolic compound stability during shelf life, within the context of the food industry. These approaches are supported by advanced chemometric tools, enabling the management and reduction of data complexity, providing effective solutions for optimizing business strategies and quality control. Regarding hazelnuts, a targeted UHPLC-HRMS approach was used to quantify bitter compounds associated with the "cimiciato" defect (such as Carpinontriol-B), providing an alternative to visual analysis for quality control. Additionally, the untargeted analysis identified molecular markers linked to the "rotten" defect, caused by bedbug attack, which were detectable even at very low concentrations (up to 1%). The thesis also focused on cocoa, aiming to develop effective tools for geographic origin discrimination. Through an untargeted approach, key molecular compounds were identified for geographic discrimination, allowing for a successful correlation of origin with specific sensory attributes using a sensomics approach. Subsequently, the GC-IMS analysis of volatile compounds, integrated with automated chemometric tools developed in Python, proposed a new rapid tool for cocoa origin verification. Lastly, the thesis addressed the stability of phenolic compounds during the shelf life of sugar-free candies enriched with botanical extracts to ensure the product’s nutritional quality during storage. In particular, the study focused on the persistence of phenolic compounds from grape seed and green tea extracts, using both targeted and untargeted UHPLC-HRMS approaches. The untargeted approach, based on phenolic fingerprinting, proved most effective for monitoring these compounds throughout the storage period.