Quantification of extra virgin olive oil in dressing and edible oil blends using the representative TMS-4,4′-desmethylsterols gas-chromatographic-normalized fingerprint

摘要: This paper describes and discusses the application of trimethylsilyl (TMS)-4,4'-desmethylsterols derivatives chromatographic fingerprints (obtained from an off-line HPLC-GC-FID system) for quantification extra virgin olive oil in commercial vinaigrettes, dressing salad in-house reference materials (i-HRM) using two different Partial Least Square-Regression (PLS-R) multivariate methods. Different data pre-processing strategies were carried out being whole one: (i) internal normalization; (ii) sampling based on The Nyquist Theorem; (iii) correlation optimized shifting, icoshift; (iv) baseline correction (v) mean centering (vi) selecting zones. first model corresponds to a matrix dimensions 'n×911' variables second one 'n×431' variables. It has be highlighted that proposed PLS-R models allow binary blends, foodstuffs, etc., when provided percentage is greater than 25%.