Raman Spectroscopic Barcode Use for Differentiation of Vegetable Oils and Determination of Their Major Fatty Acid Composition

• Published in: Journal of the American Oil Chemists' Society 2016-05, Vol.93 (5), p.627-635
• Main Authors: Velioglu, Serap Durakli, Ercioglu, Elif, Temiz, H. Tumay, Velioglu, H. Murat, Topcu, Ali, Boyaci, Ismail H.
• Format: Article
• Language: English
• Subjects: Agriculture; Biomaterials; Biotechnology; Chemistry; Chemistry and Materials Science; Chemometrics; Fatty acids; Food Science; Industrial Chemistry/Chemical Engineering; Original Paper; Principal components analysis; Raman spectroscopy; Soybeans; Spectral barcoding technique; Spectrum analysis; Studies; Vegetable oils; Vegetables
• ISSN: 0003-021X; 1558-9331
• DOI: 10.1007/s11746-016-2808-7

In this study, differentiation of vegetable oils and determination of their major fatty acid (FA) composition were performed using Raman spectral barcoding approach. Samples from seven different sources (sunflower, corn, olive, canola, mustard, soybean and palm) were analyzed using Raman spectroscopy. Second derivative of the spectral data was utilized to generate unique barcodes of oils. Chemometric analyses, namely, principal component analysis (PCA) and partial least square (PLS) methods were used for data analysis. PCA was applied for classification of the samples according to the differences in their levels arising from their barcode data. A successful differentiation based on second derivative barcodes of Raman spectra (2D-BRS) of vegetable oils was obtained. In addition, PLS method was applied on 2D-BRS in order to determine the major FA composition of these samples. Coefficient of determination values for palmitic, stearic, oleic, linoleic, α-linolenic, cis -11 eicosenoic, erucic and nervonic acids were in the range of 0.970–0.989. Limit of detection and limit of quantification values were found to be satisfactory (0.09–8.09 and 0.30–26.95 % in oil) for these fatty acids . Advantages of both chemometric analysis and spectral barcoding approach have been utilized in the present study. Taking the second derivative of the Raman spectra has minimized background variability and sensitivity to intensity fluctuations. Spectral conversion to the barcodes has further increased the quality of information obtained from Raman spectra and also made it possible to improve the visualization of the data. Converting Raman spectra of oils into barcodes enables simpler presentation of the valuable information, and still allows further analysis such as classification of vegetable oils and prediction of their major fatty acids with high accuracy.