Combined diagnostic ratio of trace elements and biomarkers with multivariate statistical analysis for the differentiation of crude oil origin

Mingxuan Zhang,Detian Yan,Tong Li
DOI: https://doi.org/10.1016/j.microc.2024.110491
IF: 5.304
2024-04-11
Microchemical Journal
Abstract:Crude oil serves as a crucial mineral resource in terms of both imports and exports, and therefore, traceability of crude oil origin is imperative. This study pioneered a novel approach to identifying crude oil origin that was specifically tailored to fraud testing in imports. The analysis included 27 representative crude oil samples from nine major exporting nations, utilizing gas chromatography–mass spectrometry (GC–MS) and inductively coupled plasma–mass spectrometry (ICP–MS). Based on these samples, a corresponding crude oil origin tracing model was established. The following significant findings were obtained: (1) There were differences in the content of trace elements, biomarkers, and diagnostic ratios among crude oil samples from different countries; (2) employing partial least squares discriminant analysis (PLS–DA), a multivariate statistical analysis model was constructed, incorporating 24 trace element and biomarker diagnostic ratios (DRs) and effectively determining crude oil origin; (3) eleven diagnostic rates with the highest discriminative power were extracted; (4) the model was tested by a confusion matrix, and the accuracy rate was 93.8%. This research method may have practical application value to better identify the origin of imported crude oil.
chemistry, analytical
What problem does this paper attempt to address?