Innovative industrial application of food metabolomics and food volatilomics for high quality peanuts

This doctoral thesis addresses the innovative industrial application of food metabolomics and volatilomics for the assessment and definition of “high quality” peanuts, with the objective of bridging advanced analytical chemistry and practical quality control strategies. The research responds to the need for objective, reliable, and transferable quality markers capable of supporting decision-making along the food production chain, where conventional approaches often fail to capture the complexity of aroma and compositional variability. The work is based on the integration of high-resolution analytical platforms, primarily comprehensive two-dimensional gas chromatography (GC×GC) coupled with mass spectrometry, enabling an in-depth characterization of the peanut volatilome. This analytical framework was further strengthened through the implementation of Untargeted–Targeted (UT) fingerprinting strategies, combining the comprehensive coverage of unknown features with the robustness of targeted compounds, and generating high-information datasets suitable for multivariate analysis and classification. A central aspect of the thesis is the application of sensomics concepts to define the molecular basis of peanut aroma quality. By integrating gas chromatography–olfactometry (GC-O) and Aroma Extract Dilution Analysis (AEDA), the study identified key odorants and established aroma blueprints, representing the minimal set of compounds responsible for the characteristic sensory profile. This approach enabled a mechanistic interpretation of aroma formation, linking volatile compounds to their precursors and associated reaction pathways, including Maillard chemistry, Strecker degradation, and lipid oxidation. Several case studies were investigated to demonstrate the applicability of the proposed approach. The chemical basis of split peanut kernels, an industrially relevant but underexplored defect, was elucidated by combining volatilome fingerprinting with primary metabolite profiling, revealing distinct compositional patterns and their impact on aroma quality. A comparative sensomics study across different geographical origins showed that aroma variability is primarily driven by differences in the relative distribution and intensity of shared odorants rather than by the presence of unique compounds, highlighting the role of precursor composition in shaping origin-dependent signatures. In addition, volatilome fingerprinting combined with multivariate chemometric modeling enabled the classification of peanuts according to industrial quality levels, while shelf-life studies identified key markers of lipid oxidation and their evolution over time. Particular emphasis was placed on quantitative volatilomics, addressing the limitations of conventional semi-quantitative approaches through the implementation of multiple headspace solid-phase microextraction (MHS-SPME), ensuring accurate, reproducible, and transferable measurements. Finally, the integration of Artificial Intelligence tools and data fusion strategies demonstrated the potential to translate complex chemical information into predictive models and decision-support systems for quality assessment, origin tracing, and process monitoring. Overall, this thesis provides a comprehensive and scalable approach for modern food quality assessment, showing how the integration of volatilomics, sensomics, and quantitative analytical strategies can significantly enhance the understanding, control, and prediction of industrial food quality.