Development of analytical and chemometric approaches for studying vegetable matrices to evaluate compositional differences and bioactive properties

This research aim to develop and apply multi-analytical and chemometric approaches to plant matrices belonging to the Brassicaceae, Asteraceae and Piperaceae families, in order to evaluate their compositional differences and bioactive properties. The study focused on the identification, extraction and characterisation of bioactive compounds, in particular polyphenols, flavonoids and glucosinolates, which are responsible for antioxidant activity and beneficial properties for human health. Non-destructive techniques such as digital microscopy (E-Eye) and attenuated total reflection infrared spectroscopy (ATR-FT-IR) were used to characterise different varieties of Brassicaceae grown in the same experimental field, eliminating soil and climate variability. SIMCA and PLS-DA chemometric algorithms enabled the classification and discrimination of local varieties, with particular attention to Mugnoli di Pettorano sul Gizio, whose hybrid nature was confirmed by multivariate analysis. A D-Optimal design was applied to optimise the extraction parameters (sample quantity, temperature, time and composition of the extracting mixture) of Brassicaceae leaves using HPLC-DAD and UV-Vis analysis. The correlation between chromatographic data and antioxidant activity was modelled using the multi-way N-PLS algorithm, obtaining a robust predictive model. Subsequently, a full factorial design was used to optimise the cold extraction of bioactive compounds from Matricaria chamomilla L. Chemometric analysis (PCA and ASCA) highlighted the effect of temperature, time and quantity on the polyphenolic profile and spectral characteristics. The optimal conditions (2.5 g, 25 °C, 32 min) produced an extract rich in hydroxybenzoic and hydroxycinnamic acids with high antioxidant activity. Finally, a multi-analytical approach was applied to Piper nigrum L. to characterise its chemical composition, antioxidant properties, and geographical traceability. The study analysed black pepper samples from Kampot (Cambodia, Protected Geographical Indication, one of which also organic), Madagascar, and commercial products, along with a green pepper sample for comparison. The aromatic profile, determined by headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry (HS-SPME-GC-MS), revealed compositional differences related to geographical origin, dominated by terpenoid compounds such as β-caryophyllene, 3-carene, limonene, and sabinene. Principal Component Analysis (PCA) enabled the identification of distinctive volatile markers associated with each provenance. Antioxidant activity, assessed through the DPPH assay, showed the highest radical-scavenging capacity in Madagascar pepper (IC₅₀ = 7.94 mg/mL), whereas Kampot pepper exhibited higher IC₅₀ values, indicating lower antioxidant efficiency. Infrared spectroscopy (ATR-FT-IR), combined with chemometric methods (PLS-DA), allowed non-destructive classification of samples according to their geographical origin, achieving an accuracy greater than 90%. Overall, the integrated approach demonstrated that combining volatile profile analysis, antioxidant evaluation, and vibrational spectroscopy provides an effective tool for the characterisation and geographical authentication of black pepper, highlighting the influence of origin and processing factors on its bioactive properties.