A Multivariate Identification of Natural Triglyceride Oils

Natural oils possess a characteristic and more or less unique pattern of fatty acids and triacylglycerols. These patterns can be used in several manners, as a means of identification, for batch-consistency testing (e. g. for the pharmaceutical and cosmetic industry), detection of adulteration and also possibly as a basis for a chemotaxonomy, revealing relationships between species. The fatty acid composition, determined by GC/MS of the methyl esters and the triacylglycerol pattern analysed by reversed phase high performance liquid chromatography, were utilized in multivariate-statistical analysis to ‘map’ and classify different oils. A class model was developed for each of the different sources of oil, constituting an objective and quantitative means of identification, through nearness in the multi-dimensional measurement space.