Multivariate assessment for predicting antioxidant activity from clove and pomegranate extracts by MCR-ALS and PLS models combined to IR spectroscopy

[Display omitted] •MCR-ALS coupled to NIR spectroscopy efficiently determined AA% of clove and pomegranate extracts.•Phenolic compounds with high AA% were identified in the naturals extracts by NMR.•A methodology combining MIR/NIR spectroscopy and MCR-ALS and PLS was proposed.•A low-cost multivariate approach alternative to DPPH standard method for AA% was reached. This study evaluated the feasibility of infrared (MIR/NIR) spectroscopy coupled to chemometrics as an alternative method for determining the antioxidant activity (AA%) of pomegranate (Punica granatum) and clove (Syzygium aromaticum) alcoholic extracts versus the conventional DPPH method. Multivariate curve resolution with alternating least squares (MCR-ALS) and Partial least squares (PLS) regression were efficient to predict the AA%, thus providing good accuracy and low residuals compared to the standard method. The MCR-ALS combined with NIR data stood out among the other models with R2 ≥ 0.962 and RMSEP ≤ 3.38 %; furthermore, this technique presents the great feature of recovering the pure spectral profile of the analytes and identifying interferents in the sample. The application of chemometrics tools to predict the antioxidant activity of natural extracts resulted in a greener, low-cost and efficient process for the food industry.