Mid-infrared spectroscopy using Fourier transform infrared (FTIR) with attenuated total reflectance (ATR) correction was coupled with partial least square regression (PLSR) for the prediction of phenolic acids and flavonoids in fig (peel and pulp) identified with high-performance liquid chromatography-diode array detector (HPLC-DAD), with regards to their partitioning between peel and pulp. HPLC-DAD was used to quantify the phenolic compounds (PCs). The FTIR spectra were collected between 4,000 and 450 cm-1 and the data in the wavenumber range of 1.175-940 cm-1, where the deformations of O-H, C-O, C-H, and C=C corresponded to flavanol and phenols, were used for the establishment of PLSR models. Nine PLSR models were constructed for peel samples, while six were built for pulp extracts. The results showed a high-throughput accuracy of such an approach to predict the PCs in the powder samples. Significant differences were detected between the models built for the two fruit parts. Thus, for both peel and pulp extracts, the coefficient of determination (R2) ranged from 0.92 to 0.99 and between 0.85 and 0.95 for calibration and cross-validation, respectively, along with a root mean square error (RMSE) values in the range of 0.46-0.9 and 0.23-2.05, respectively. Residual predictive deviation (RPD) values were generally satisfactory, where cyanidin-3,5-diglucoside and cyanidin-3-O-rutinoside had the higher level (RPD > 2.5). Similar differences were observed based on the distribution revealed by partial least squares discriminant analysis (PLS-DA), which showed a remarkable overlapping in the distribution of the samples, which was intense in the pulp extracts. This study suggests the use of FTIR-ATR as a rapid and accurate method for PCs assessment in fresh fig.
Keywords: FTIR-ATR; Ficus carica L.; HPLC-DAD; partial least square regression (PLSR); phenols.
Copyright © 2022 Hssaini, Razouk and Bouslihim.