Pacific and Yukon Laboratory for Environmental Testing, Science and Technology Branch, Pacific Environmental Science Centre, Environment and Climate Change Canada, North Vancouver, British Columbia, V7H 1B1, Canada.
Agilent Technologies Inc., Santa Clara, CA, USA.
J Chromatogr A. 2020 Dec 20;1634:461689. doi: 10.1016/j.chroma.2020.461689. Epub 2020 Nov 11.
Artificially weathered crude oil "spill" samples were matched to unweathered suspect "source" oils through a three-tiered approach as follows: Tier 1 gas chromatography-flame ionization detection (GC/FID), Tier 2 gas chromatography-mass spectrometry (GC/MS) diagnostic ratios, and Tier 3 multivariate statistics. This study served as proof of concept for a promising and new method of crude oil forensics that applies principal component analysis (PCA) and partial least squares discriminant analysis (PLSDA) in tandem with traditional forensic oil fingerprinting tools to confer additional confidence in challenging oil spill cases. In this study, weathering resulted in physical and chemical changes to the spilled oils, thereby decreasing the reliability of GC/FID and GC/MS diagnostic ratios in source attribution. The shortcomings of these traditional methods were overcome by applying multivariate statistical tools that enabled accurate characterization of the crude oil spill samples in an efficient and defensible manner.
通过以下三个层次的方法，将人工风化的原油“泄漏”样本与未风化的可疑“源”油相匹配：第 1 层气相色谱-火焰离子化检测（GC/FID）、第 2 层气相色谱-质谱（GC/MS）诊断比和第 3 层多元统计。本研究证明了一种有前途的新的原油取证方法的概念，该方法将主成分分析（PCA）和偏最小二乘判别分析（PLSDA）与传统的法医油指纹识别工具相结合，为具有挑战性的溢油案件提供了额外的信心。在本研究中，风化导致了溢油的物理和化学变化，从而降低了 GC/FID 和 GC/MS 诊断比在源归属中的可靠性。通过应用多元统计工具克服了这些传统方法的缺点，这些工具能够以有效和有防御性的方式准确地描述原油溢油样本。
