Application of Near Infra Red (NIR) Spectroscopy and Chemometrics for Determination Antioxidant Activity of Plant Leaves Extracts

Abstract

Context: Near Infra red (NIR) spectroscopy combined with chemometrics has been developed for simple analysis of antioxidant activity in the medicinal plant extract. Aims: The aim of this research was to study whether NIR and chemometric methods could be used to determine the antioxidant activity. Settings and Design: A model for determining antioxidant activity of plant extracts was formed using chemometric and the model was applied on real samples. Methods and Material: Medicinal plant leaves were extracted and its spectral data were correlated with its antioxidant activity using chemometric. The chemometric method used for calibration analysis were Partial Last Square (PLS), Principal Component Regression (PCR) and Support Vector Machines Regression (SVMR), and the methods used for classification analysis were Linear Discriminant Analysis (LDA), Soft Independent Modelling of Class Analogies (SIMCA), and Support Vector Machines Classification (SVMC).Statistical analysis used: Paired–sample t test was used in this study. Results: In this study, SVMC that showed best classification with accuracy was 100 % and SVMR that showed best calibration with R2 and RMSEC value was 0.9 811 205 and 4.4 940 028, respectively. SVMC and SVMR models were further used to predict unknown antioxidant activity in commercial and simulation samples. Using these models, the significance of antioxidant activity that has been measured by NIR and UV‒Vis spectrophotometry was evaluated with paired samples ttest and gave no significant difference. Conclusions: This study showed that NIR spectroscopy coupled with chemometric could be used for determining antioxidant activity of several plant extracts.