Classification of vegetable oils according to their botanical origin using n-alkane profiles established by GC–MS

• Published in: Food chemistry Vol. 167; pp. 36 - 39
• Main Authors: Troya, F., Lerma-García, M.J., Herrero-Martínez, J.M., Simó-Alfonso, E.F.
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 15.01.2015; Elsevier
• Subjects: Alkanes - chemistry; Biological and medical sciences; Botanical origin; Food toxicology; Gas chromatography – mass spectrometry; Gas Chromatography-Mass Spectrometry - methods; Linear discriminant analysis; Medical sciences; n-Alkanes; Plant Oils - chemistry; Toxicology; Vegetable oils; Vegetable oils; n-Alkanes; Linear discriminant analysis; Botanical origin; Gas chromatography – mass spectrometry; Origin; Discriminant analysis; Vegetable oil; Profile; Gas chromatography - mass spectrometry; Gas chromatography; Alkane; Classification; Mass spectrometry
• ISSN: 0308-8146; 1873-7072
• DOI: 10.1016/j.foodchem.2014.06.116

•An LDA model for vegetable oil classification according to their botanical origin was constructed.•n-Alkane profiles established by GC–MS were used for classification.•An excellent resolution among the different vegetable oils was achieved. n-Alkane profiles established by gas chromatography-mass spectrometry (GC–MS) were used to classify vegetable oils according to their botanical origin. The n-alkanes present in corn, grapeseed, hazelnut, olive, peanut and sunflower oils were isolated by means of alkaline hydrolysis followed by silica gel column chromatography of the unsaponifiable fractions. The n-alkane fraction was constituted mainly of n-alkanes in the range C8–C35, although only those most abundant (15 n-alkanes, from 21 to 35 carbon No.) were used as original variables to construct linear discriminant analysis (LDA) models. Ratios of the peak areas selected by pairs were used as predictors. All the oils were correctly classified according to their botanical origin, with assignment probabilities higher than 95%, using an LDA model.