The composition and tectonic setting of Earth's earliest crust are fundamental questions in Earth science, but are difficult to determine due to the scarcity of sufficiently old rocks. Previous research on 4.4–3.3 Gyr old zircons from the Jack Hills in Western Australia has yielded conflicting interpretations regarding their protoliths. Some studies suggest low-temperature, hydrous granites, while others propose more mafic, intra-plate basaltic sources or impact melts. These differing interpretations have implications for understanding the early operation of plate tectonics, with some models suggesting a stop-start mode. Many conclusions are drawn from Hf isotope signatures, therefore a complementary approach using trace element data is needed. Previous analyses of rare earth element (REE) abundances in Jack Hills zircons have indicated a continental igneous origin, but further information can be obtained from a wider range of trace elements. This study aims to extract additional information by analyzing these trace element data to determine the protolith composition and tectonic setting.

The debate surrounding the origin and composition of Earth's early crust is extensive. Studies focusing on Jack Hills zircons have yielded diverse interpretations. Some researchers argue for a low-temperature, hydrous granite protolith, suggesting early Earth conditions similar to today. Others propose a more mafic, possibly intra-plate basaltic source, or even impact melts as the origin of these zircons. These different viewpoints have significant implications for understanding the timing and nature of plate tectonics during the Hadean and early Archaean eons. While Hf isotopes have provided some insights, a complementary approach using trace elements is necessary to resolve these conflicting interpretations. Previous work has already suggested a continental igneous origin for the zircons, but a more comprehensive analysis of trace element data offers the potential for a more refined understanding of the protolith's composition and tectonic setting.

This study analyzed a new aliquot of Jack Hills zircons. U-Pb ages were determined using standard SHRIMP methodology, ranging from 4.3 to 3.3 Ga. Trace element data were obtained from the same locations using LA-ICP-MS. Zircon/melt partition coefficients were crucial to calculate the trace element composition of the melts (protoliths). The study carefully selected partition coefficients from an experimental study that analyzed a wide range of trace elements consistent with the lattice strain model, focusing on an andesitic composition that proved to be highly applicable to the Jack Hills zircons. The authors acknowledge the temperature dependence on absolute partition coefficients but note that ratios of partition coefficients are less sensitive to this. The study addressed potential inaccuracies in empirical partition coefficients arising from micro-inclusions and heterogeneous host rocks. The SiO2 content of the protolith melts was calculated using a regression of GEOROC database analyses, considering statistical and analytical errors.

The study's key findings are: 1. The average SiO2 content of the melts from which the Jack Hills zircons crystallized was 59 ± 6 wt.%. 2. The Th/Nb, Dy/Yb, and Sr/Y ratios of these melts were 2.7 ± 1.9, 0.9 ± 0.2, and 1.6 ± 0.7, respectively. 3. These ratios, coupled with the SiO2 content, strongly suggest that the protoliths were andesites formed in modern subduction settings. 4. There was no statistically significant secular change in these parameters between 4.3 and 3.3 Gyr. The calculated melt compositions, with their elevated SiO2 contents and Th/Nb ratios, are characteristic of andesitic melts in modern subduction settings. The data clearly distinguishes Jack Hills protoliths from oceanic rocks, TTGs, and Sudbury impact melts. Even when using different zircon/melt partition coefficients, the fundamental conclusion of an andesitic protolith remains consistent. The inferred protoliths had flat REE patterns and low Dy/Yb ratios, further distinguishing them from TTGs. The average composition of the inferred protoliths is strikingly similar to low-Ti enriched basaltic andesites from the Nuvvuagittuq Greenstone Belt, which are thought to originate from a subduction initiation sequence. The lack of within-plate geochemical signatures or alternation between within-plate and subduction-like signatures over time is also a significant finding.

The study's findings strongly suggest that the Jack Hills zircons crystallized from melts with geochemical signatures typical of modern arc andesites. This is consistent with the early operation of modern-style plate tectonics. The similarity between the inferred protolith composition and the Nuvvuagittuq Greenstone Belt andesites further supports this interpretation, linking the early Hadean to later Archaean crustal evolution processes associated with subduction. The absence of evidence for within-plate signatures throughout the studied time period contradicts models proposing a stop-start mode for early plate tectonics. This aligns with recent geodynamic modelling and isotopic studies suggesting a very early onset of plate tectonics on Earth.

This study provides strong evidence that the protoliths for the Jack Hills zircons were andesites formed in a subduction setting, suggesting that plate tectonics may have been active very early in Earth's history. The consistency of this finding across the studied time interval implies a remarkably stable tectonic environment. Future research could focus on expanding the dataset to include a broader range of Hadean and early Archaean zircon populations to further test this hypothesis and constrain the timing and nature of early plate tectonics.

The study acknowledges the reliance on zircon/melt partition coefficients, highlighting the potential influence of coefficient selection on the calculated melt compositions. While the authors chose coefficients consistent with the lattice strain model and deemed suitable for the andesitic protolith composition, the use of different coefficients could alter the interpretation. Also, although this study examined a new aliquot of zircons, the interpretation is ultimately still based on a relatively small sample of zircons from a single locality.