Authentication of avocado oil mixed with cooking oils (branded and loose palm oil) utilizing fourier transform infrared spectroscopy in conjunction with chemometrics

Abstract

Avocado oil (AVO) has become popular for its health advantages and culinary variety, while its elevated cost has rendered it susceptible to adulteration practices, frequently with less expensive edible oils such as palm oil. Attenuated total reflectance-fourier transform infrared spectroscopy (ATR-FTIR) in conjunction with chemometrics is a promising way to quickly, rapidly, and non-destructively detect AVO adulterated with branded palm oil (BPO) or loose palm oil (LPO). AVO can be distinctly categorized apart from BPO and LPO using FTIR analysis corroborated by principal component analysis (PCA). Principal component (PC) 1 is primarily influenced by absorption at a wavenumber of 1743 cm⁻¹, whereas PC2 is impacted by absorption at a wavenumber of 2953 cm⁻¹. With first derivative data and multiplicative scatter correction (MSC), multivariate calibration of principal component regression (PCR) was able to correctly predict the amount of BPO in AVO. This approach yielded root mean squared error of calibration (RMSEC) and root mean square error for prediction (RMSEP) values of 0.56 and 0.55, alongside coefficient of determination for the calibration and prediction model (R²-cal and R²-val) values of 0.9998 and 0.9999, respectively. In addition, PCR using first-derived data without MSC adjustment accurately quantified BPO levels in AVO, with RMSEC and RMSEP values of 0.60 and 0.61, respectively, along with R²-cal and R²-val values of 0.9998. This PCA-based discriminant analysis correctly separated the real AVO samples from the adulterated AVO samples with classification accuracies of 98.18 % for BPO and 100 % for LPO. This method may also be applicable further for detecting other possible edible oil adulteration, including soybean oil, rapeseed oil, and sunflower oil.