Construction Of Triacylglycerols Libraries For Part Of The Edible Vegatable Oils And Its Application In Detecting Adulteration

Edible vegetable oil is the essential nutrients for human beings. Shoddy and realistic-looking edible vegetable oil can be found everywhere around the market. At present, the detection technology developed is still cannot identify adulterated edible vegetable oil. In order to solve this problem, reversed phase liquid chromatography with atmospheric pressure chemical ionization source tandem ion trap has been employed for analysis of triacylglycerols in different vegetable oil, and chemometrics method were used:the principal component analysis, random forest, clustering analysis. Relevant models were built for accurately differentiated varieties of edible vegetable oil using partial least squares, and they have and achieved to identify adulterated edible vegetable oil. The main study contents were summarized as follows:(1) The structure identification technique for triacylglycerols was eastablished by reversed-phase liquid chromatography with atmospheric pressure chemical ionization source-ion trap mass spectrometry. The collision energy and ionization way was optimized by 1-palmitoyl-2-Oleoyl-3-stearoyl-rac-glycerols (POS) as the research object.. Results show that stable fragment ions can be produced by the collision energy of 25-80 eV with atmospheric pressure chemical ionization source.(2) Construction of triacylglycerols libraries in edible vegetable oils, triacylglycerols were analyzed for the rapeseed oil, soybean oil, tomato seed oil, olive oil, melon seed oil, walnut oil, safflower seed oil, peanut oil, rice bran oil, cottonseed oil, pumpkin seed oil, grape seed oil, pomegranate seed oil, sesame oil, sunflower seed oil, tea seed oil, milk thistle oil, almond oil, flax seed oil, corn oil and perilla seed oil using the established structure identification technology. Total ion chromatorgraphy library and APCI ion source mass spectrometry ions database have been build for 21 kinds of edible vegetable oils. The triacylglycerols’ types and contents for each vegetable oil have been compared, and the characteristic triacylglycerols in some oils were obtained.(3) Establishment of adulterated identification technique for edible vegetable oil based on triacylglycerols libraries. Peanut oil, sesame oil, soybean oil, sunflower seed oil and flax seed oil as target oils, models of principal component analysis (PCA) and random forest (RF) were established base on triacylglycerols libraries. Five edible vegetable oils were distinguished using PCA and RF. The analysis result for principal component analysis model has shown that LLL、OLL、OOL、PLL、SOL、POL are the most important factors. The flax seed oil and no flax seed oil were successfully distinguished using the clustering analysis. LnLnLn, LnLLn, OLnLn, PLnLn are the most important factors. The adulterated identification of edible vegetable oils were quantified using partial least squares discriminant analysis model. The prediction results showed that flax seed oil mixed with more than 5% mass concentration of cheap oil could be successful indentifiedby partial least squares discriminant analysis model. The varieties of edible vegetable oil in flax blend oil can be qualitative identified by random forests modle and that accuracy was obtained as high as 90.46%.