Determinaton of some quality parameters of oi̇ls and fats by using NIR and reference methods

In this study, the peroxide value and free fatty acid content various edible oils such as sunflower oil, corn oil, virgin and riviera olive oil and animal fats such as butter and chicken fats were compared using reference methods and NIR method. Other quality charecteristics of edible oils and fats including tocopherol, phenol, chlorophyll, carotenoid and malonaladehyde contents and colour values of oils (L, a, b and Hue) were determined by using reference methods The peroxide value (PV) and free fatty acidity (FFA) determinede by both methods were very close and a high level of relationship was found between them determide by the chemometric methosd including PLS (Pricipal Least Square-Regression) (R²=0.93, 0.99, respectively) and PCA (Principal Component analysis. The statistical differences between PV and FFA values determined by both reference and NIR methods were analysed by using two component t-test at levels of p<0.05 and 0.01. Oils and fats were classified according to the resulst of FFA and PV obtained by both method and evaluated by PCA method. Acording to PCA score graphics it was found that the extra virgin olive oil that have PV values ranging betwen 7.7 and 26.7 meq O2/kg oil were completely seperated from other edible oils and from chicken fat ((R²=0.93, p<0.05). The reslts obtained for FFA contnet of edible oils and fats by both methods, first principal component had 99.79 % and second principal component comprises of 0.21 % of total variance. It was found that, according to PCA score graphics, FFA content of chicken fat ranged between 6.6 to 12.1 % (as oleic acid) and seperated completely from other edible oils (R²=0.99, p<0.05). Edible oils and fats are classified according to quality criteria using PCA data. According to PCA score graphics of chlorohyll, carotenoid, L, a, b and Hue values, firts principal component had 48.1% and second principal component had 39.7 % of total varieance. It was found that, according to colour values butter and chicken fat were sperated from other edible oils and were in completely in different groups. The resulsts of this study clearly indicated that some quality charecteristics of edible oils and fats could be determined by NIR method with multivariate statistical analysis such as PLS-R and PCA in a fast, reliable and simultaneous way. The NIR method valuable analytical tool fort he quality control and classification of edible oils and fats.