Valorisation of Mangrove Leaf Powders: Physicochemical Characterisation and FT-MIR Monitoring of Linseed Oil Stability

This study investigated Fourier-transform mid-infrared (FT-MIR) spectroscopy coupled with chemometric tools as a rapid, non-destructive approach for monitoring oxidation in linseed oil supplemented with mangrove-derived extracts. Young and mature leaves of Avicennia marina and Rhizophora mucronata were characterised for water activity, moisture, pH, protein, ash, mineral composition, and colour, providing a comparative raw-material profile before extract preparation. The corresponding extracts were evaluated in linseed oil during a 15-day Schaal oven test, alongside butylated hydroxytoluene (BHT), rosemary extract, and an untreated control. FT-MIR spectroscopy was used to follow changes in three spectral regions: 1200–1150 cm−1, associated with the C–O/ester-related band; 1742–1720 cm−1, linked with the carbonyl interval; and 1532–1510 cm−1, related to aromatic compounds. Chemometric modelling was intentionally restricted to the 1200–1150 and 1742–1720 cm−1 intervals to assess storage-related variation in the C–O/ester and carbonyl features of the oil matrix. Factorial discriminant analysis (FDA) applied to the first five principal components of the principal component analysis (PCA) of each spectra region achieved leave-one-spectrum-out classification accuracies of 58.10% and 71.43% for the C–O/ester-related and carbonyl models, respectively. The FDA applied to the concatenated both oxidation-related regions improving classification to 76.19%. The combined model captured storage-related spectral evolution more effectively than either individual region or revealed treatment-dependent trajectories relative to untreated oil. These findings complemented conventional oxidation indices and show that targeted FT-MIR spectroscopy coupled with chemometric tools provide a practical framework for monitoring oxidative changes and screening plant-derived preservation strategies in edible-oil systems.