Trace element tracing of oil-source rock correlations in petroliferous basins is a cutting-edge area of research in the field of petroleum geochemistry, providing an important supplement for the traditionally used organic geochemistry approach. However, the specific proxies and mechanisms are not well understood, limiting the application of the trace element approach and theoretical geochemical behavior during the oil generation, migration and accumulation. To fill the knowledge gap, based on a case study in the lower Permian Fengcheng Formation alkaline-lacustrine petroleum system, located in the northwestern Junggar Basin, Northwest China, novel trace element proxies for oil-source rock correlations was evaluated with the geochemical behavior and mechanisms being explored. Trace elements and conventional geochemistry in twenty source rock and nineteen crude oil samples were analyzed by inductively coupled plasma-mass spectrometry (ICP-MS) and gas chromatography-mass spectrometry (GC-MS) to systematically compare. Analysis of 33 commonly-detected trace elements showed that the transition metals (Ti, V, Cr, Ni, Cu, Zn, Mn, and Zr), alkaline earth metals (Sr and Ba), and nonmetallic elements (B and As) exhibit good correlations between each source rock and crude oil. However, the contents of these 12 elements exhibit changes by a factor of 0.3-1141 times from source rock to crude oil. Elements Ni, Cu, Zn and Cr are relatively enriched in crude oils, elements V, Ti, Mn and Zr are relatively enriched in source rocks, and elements B, As, Sr and Ba are not apparently enriched. This indicates that the elements contents cannot be directly used for oil-source rock correlations. Understanding the geochemical behavior and mechanisms of the differential enrichment of trace elements from source rock to crude oil is prerequisite. It is argued that the differential enrichment is mainly controlled by the geochemical partition of these elements, in the form of organic complexes (and chelates) or inorganic ions during hydrocarbon generation and accumulation. In detail, Ni and some V are chelated to organic porphyrin compounds by strong N-Ni/V bonds, which are little affected by petroleum migration and are thus relatively stable. In addition, most of the V occurs in insoluble mineral phases, and thus is preferentially enriched in source rocks during primary hydrocarbon migration. Chromium, Cu, and Zn are mainly complexed with O or S atoms in organic acids in the form of Cr-O or Cu (Zn)-S bonds. These chemical bonds are readily broken during secondary petroleum migration. Manganese, Zr, Ti, Sr, and Ba occur mainly in mineral phases in the source rocks and as ions during petroleum migration, and thus are modified by oil-rock interactions. Boron occurs mainly in mineral phases in the source rocks, but is easily complexed with organic acids, and thus undergoes limited modification due to secondary hydrocarbon migration. Arsenic occurs mainly as arsenide or arsenate ions, which are strongly affected by secondary petroleum migration. As such, Ni, V, B, Cr, Cu, Zn, and As are more reliable for oil-source rock correlations and are incorporated into four new parameters established in this study: (1) Ni contents can be used for oil-source rock correlations; (2) V/Ni and Ni/B ratios can be used to trace primary petroleum migration; and (3) a Cr-(Cu+Zn)-As ternary diagram can be used to trace secondary petroleum migration. In conclusion, trace elements in petroleum systems exhibit complex geochemical behavior during hydrocarbon generation and primary (secondary) oil migration. However, robust trace element proxies for oil-source rock correlations can be established after consideration of the geochemical behavior of each element. The proposed method is an important addition to conventional oil-source rock correlations based on organic geochemistry.
