As a proxy of magmatic oxidation state, the accurate characterization of the Ce anomaly of zircon is of great significance since it can give important information for provenance studies of rocks as well as for exploration of intrusion-related mineral deposits. The magnitude of the zircon Ce anomaly has been traditionally described by Ce/Ce*, where Ce* is the theoretical Ce value derived from a chondrite-normalized rare earth element (REE) pattern. More recently, the Ce⁴⁺/Ce³⁺ method based on the lattice strain model has been proposed, since the latter method does not need La and Pr contents for zircon, both of which are commonly below the limit of detection and susceptible to contamination from melt/mineral inclusions. In this contribution we show that the Ce⁴⁺/Ce³⁺ method is confronted with some problems in practice and should be further improved. In contrast, by re-examining chondrite-normalized REE patterns of zircon, we find that Ce* can be estimated according to a logarithmic function curve without involvement of La and Pr contents. Application of this new method to zircon data from 11 giant to supergiant porphyry Cu deposits suggests this revised method as a more valid measure in evaluating magmatic oxidation state. The revised Ce/Ce* method is of particular importance for analyses where the provenance of the analyzed zircon is unknown or in question, since the method does not require knowledge of the melt composition.

The Weibao copper–lead–zinc skarn deposit is located in the northern East Kunlun terrane, NW China. Igneous intrusions in this deposit consist of barren diorite porphyry (U–Pb zircon age of 232.0 ± 2.0 Ma) and ore-bearing quartz diorite and pyroxene diorite (U–Pb zircon ages of 223.3 ± 1.5 and 224.6 ± 2.9 Ma, respectively). Whole-rock major and trace element and accessory mineral (zircon and apatite) composition from these intrusions are studied to examine the different geochemical characteristics of ore-bearing and barren intrusions. Compared to the barren diorite porphyry, the ore-bearing intrusions have higher Ce⁴⁺/Ce³⁺ ratios of zircon and lower Mn contents of apatite, indicating higher oxidation state. Besides, apatite from the ore-bearing intrusions shows higher Cl contents and lower F/Cl ratios. These characteristics collectively suggest the higher productivity of ore-bearing quartz diorite and pyroxene diorite. When compared with ore-bearing intrusions from global porphyry Cu deposits, those from Cu–Pb–Zn skarn deposits display lower Ce⁴⁺/Ce³⁺ and Eu_N/Eu_N* ratios of zircon and lower Cl and higher F/Cl ratios of apatite. We conclude that these differences reflect a general geochemical feature, and that zircon and apatite composition is a sensitive tool to infer economic potential of magmas and the resulting mineralization types in intrusion-related exploration targets.

Zircon geochemistry provides a sensitive monitor of its parental magma composition. However, due to the complexity of the uptake of trace elements during zircon growth, identifying source magmas remains challenging, particularly for detrital grains whose petrological context is lost. We use a machine learning-based approach to explore the classifiers for zircon provenance, based on 3794 published, high-quality zircon trace element analyses compiled from I-, S-, and A-type granites. Three supervised machine learning algorithms, namely, Support Vector Machine (SVM), Random Forest (RF), and Multilayer Perceptron (MLP) were used and trained with 11 features, including 7 trace elements (Ce, Eu, Ho, Nb, Ta, Th, and U) and 4 derived trace element ratios (Th/U, U/Yb, Ce/Ce*, and Eu/Eu*). Our results show that all three trained machine learning methods perform very well with accuracy varying from 0.86 to 0.89, and that input–output relationships captured by different ML methods are nearly consistent and can be explained by the known petrological processes. The application of our trained machine learning classifiers to detrital zircon studies will enhance the interpretability of zircon assemblages of different origins. It also helps develop interpretations, approaches, and tools that will benefit, for example, the study of continental crust evolution and mineral exploration.

The composition and origin of Earth’s earliest continental crust remains enigmatic due to the absence of Hadean (>4 Ga) age rocks. Here we address this question by using machine learning to examine the provenance of the 4.4–3.3 Ga Jack Hills zircons, which constitute the best archive of Earth’s earliest continental crust. Our results reveal that although some Jack Hills zircons may be derived from trondhjemite-tonalite-granodiorite series rocks, which were common during the Archean (4–2.5 Ga), most (as high as ~70%) are sourced from igneous (I-) and sedimentary (S-) type granites. This finding provides clear evidence for rocks other than the trondhjemite-tonalite-granodiorite suite in the Earth’s earliest continental crust. Considering that I- and S-type granites are typical of modern convergent plate margins, the presence of a high proportion of Jack Hills zircons from these rocks supports the operation of a horizontal, mobile-lid tectonic regime in the early Earth.