Detection of Argan Oil (Argania Spinosa L.) Adulteration with Avocado, Flaxseed, Walnut, and Pumpkin Seed Oils Using Transform Infrared Spectroscopy (FTIR) and Advanced Chemometric and Machine Learning Techniques

An increasing trend in the popularity of organic and natural products can leave them prone to adulteration. Argan Oil (AGO), derived from the kernels found in the argan nut [Argania spinosa (L.)], is a multi-beneficial health and cosmetic product that has shown to be in increasing demand. This rise in demand also leads to the increasing tendency for AGO to be adulterated with low grade or less expensive alternatives. Therefore, the overall goal of this study was to utilize a Fourier transform infrared spectroscopic and chemometric methods such as partial least squares (PLS), principal component regression (PCR), and artificial neural network (ANN) for the authentication of AGO. To achieve this, AGO was first adulterated with avocado oil (AVO), pumpkin seed oil (PSO), flaxseed oil (FSO) and walnut seed oil (WNO) in varying concentrations. PLS was able to detect AGO at a higher concentration in the training set compared to PCR and ANN, however, PCR and ANN detected a higher concentration of AGO in the testing sets. When comparing the three chemometric methods used in this study, the PCR model predicted the most desirable R2 (closest value to 1) however had the highest RMSEP values for AGO with an R2 of 0.7777 and RMSEP of 18.461 in the testing sets. All methods were able to detect the FSO adulterant at the highest concentration across all models with the highest R2 value of 0.9723 detected in the PLS model and the lowest R2 value of 0.9611 detected in the ANN model of the training sets. The WNO adulterant was detected at the lowest concentration in all the models with its highest and lowest R2 values of 0.5835 and 0.018 in the PLS and ANN models respectively. The developed ANN or PCR model could possibly be applied for the rapid detection of vegetable oil adulterants in AGO.