Rapid Discrimination Between Authentic and Adulterated Andiroba Oil Using FTIR-HATR Spectroscopy and Random Forest

Abstract

Andiroba oil (Carapa guianensis Aubl) from the Amazonian region is one of the most important oils in popular medicine and cosmetic industry, has a high added value and is liable to adulteration with other cheaper or lower quality oils. Based on these facts, this work proposes a simple, fast, and accurate methodology to discriminate authentic andiroba oil from adulterated andiroba oil-containing soybean and corn oils in different proportions. The approach is based on the applicability of FTIR-HATR spectroscopy combined with chemometrics tools for classification such as random forest and partial least squares for discriminant analysis (PLS-DA). The random forest model showed 100% of correct classification and, consequently, an accuracy of 100% in training and test sets, while PLS-DA presented an accuracy of approximately 100 and 94% in training and test sets, respectively. Therefore, the developed methodology can be helpful in routine laboratories, regulatory agencies, and industry for quality analysis of andiroba oil.