Quantitative models for prediction of palm olein adulteration in coconut testa oil

Abstract

This study was carried out to produce quantitative models for prediction of palm olein (PO) adulterations in coconut testa oil (CTO) by using fatty acid compositional analysis and differential scanning calorimetric (DSC) analysis. Samples of authentic CTO were prepared using mature coconuts collected out of three local cultivars while a pure sample of PO was obtained from a reliable supplier. In this experiment, samples of CTO were blended with PO in the range of 20 to 80% (w/w) and subjected to fatty acid and thermal analysis using the relevant instruments. Fatty acid data and DSC data of thermal transitions namely, peak temperature, peak area, peak onset and peak endset temperatures were employed to perform statistical analysis. Results showed that all fatty acids were suitable as parameters to develop quantitative models for prediction of adulteration in CTO. The highest positive correlation was displayed by linoleic acid (+0.977; p<0.0001) followed by oleic (+0.973; p<0.0001), palmitic (+0.972; p<0.0001) and stearic (+0.959; p<0.0001) acids. Out of all, the two models formed using palmitic and oleic acids were the best for prediction. According to DSC analysis, the highest correlations were found for peak maxima (-0.965; p<0.0001) and peak area (-0.951; p<0.0001). Out of these two, the model formed using peak maxima was the best for prediction.