Prolonged history of episodic fluid flow in giant hematite ore bodies: Evidence from in situ U–Pb geochronology of hydrothermal xenotime

Absolute ages for hydrothermal mineralization and fluid flow are critical for understanding the geological processes that concentrate metals in the Earthʹs crust, yet many ore deposits remain undated because suitable mineral chronometers have not been found. The origin of giant hematite ore deposits, which are hosted in Precambrian banded-iron formations (BIFs), remains contentious. Several models have been formulated based on different sources and timing for the mineralizing fluids; supergene-metamorphic, syn-orogenic, late-orogenic extensional collapse and syn-extensional. Precise geochronology of the ore offers a means of discriminating between these models. In this study, two U–Pb chronometers, xenotime and monazite, have been identified in high-grade hematite ore bodies from the Mount Tom Price mine in the Hamersley Province, northwestern Australia. Both phosphate minerals occur as inclusions within the hematite ore and as coarser crystals intergrown with martite (hematite pseudomorph after magnetite) and microplaty hematite, indicating that the xenotime and monazite precipitated during mineralization. In situ U–Pb dating by ion microprobe indicates that both phosphate minerals grew during multiple discrete events. Our results suggest that ore genesis may have commenced as early as ∼ 2.15 Ga, with subsequent hydrothermal remobilization and/or mineralization at ∼ 2.05 Ga, ∼ 1.84 Ga, ∼ 1.67 Ga, ∼ 1.59 Ga, ∼ 1.54 Ga, ∼ 1.48 Ga and ∼ 0.85 Ga. The location of the ore bodies along ancient fault systems, and the coincidence of at least some of the U–Pb phosphate dates with episodes of tectonothermal activity in the adjacent Proterozoic Capricorn Orogen, implies that fluids were channelled through major structures in the southern Pilbara Craton during discrete phases of tectonic compression and extension. Our results show that the hematite ore bodies formed at sites of repeated focussed hydrothermal fluid flow. In contrast to the aforementioned models, our results imply that iron-ore formation was probably a long-lived, multi-stage process spanning more than one billion years.