Garnet, zircon and rutile U-Pb systematics of eclogite xenoliths from the Navajo Volcanic Field (USA)

Intrusions of the Navajo Volcanic Field (NVF) contain eclogite xenoliths that record processes related to the subduction of the Farallon plate beneath the Colorado Plateau. Previous geochronological work sparked controversies about their origin, especially whether they are derived from oceanic crust of the Farallon plate, or from older continental lithosphere, based on occasional Proterozoic zircon U-Pb ages. Moreover, the mechanisms and timescales of the recorded high-pressure processes, including several stages of fluid metasomatism, are largely unknown. We study the U-Pb systematics of garnet, zircon and rutile by LA-ICP-MS in order to achieve a refined petrochronologic interpretation of the NVF eclogites.The eclogite xenoliths are hosted by serpentinized ultramafic microbreccia (SUM) which intruded the Colorado Plateau at ~30 Ma as a consequence of extensive hydration of the lithospheric mantle by Farallon slab-derived fluids. In contrast to kimberlite-borne eclogite xenoliths, which often contain garnet and omphacite only, those of the NVF additionally contain ubiquitous rutile, and often pyrite, phengite, zoisite pseudomorphs after lawsonite (with rare lawsonite relics), accessory monazite, and rare coesite. Based on this assemblage, peak P-T conditions around 4 GPa and 600C are estimated. Subsequent rapid uplift in the sub-solidus SUM is not thought to have caused significant further heating. Except for a few instances where a mid-ocean ridge basalt-like bulk chemical composition is essentially preserved, the xenoliths are strongly overprinted by several metasomatic events in the eclogite facies. Most notably, this involved interaction with a Na-Si-S-rich fluid, probably of crustal origin, and a later (just prior to exhumation) serpentinite-derived fluid.All our U-Pb rutile data as well as published U-Pb monazite data (~29 Ma) agree with the ~30 Ma SUM formation age. The majority of the U-Pb zircon analyses predate the rutile data by not more than several Myr, with a minority of older ages forming a continuum to the late Cretaceous. Unlike some earlier studies, we did not obtain any Proterozoic U-Pb zircon ages. The garnet U-Pb dates, which are mostly from pre-metasomatic zones, partly agree with the zircon dates within uncertainty, but sometimes predate the zircon dates from the same sample by up to several tens of Myr.Despite moderate peak metamorphic temperatures (~600C), rutile remained an open system for Pb until exhumation. Garnet and zircon are resistant to Pb volume diffusion at these temperatures, however, the zircon ages appear to be largely reset during fluid metasomatism, as also indicated by often high common Pb contents. Due to the absence of preserved igneous zircon ages, the protolith origin remains uncertain. Garnet, which grew mostly before metasomatism, seems to provide robust ages of initial eclogitization, which may have been diachronous in our sample suite. Geochronological evidence implies that several tens of Myr passed between initial eclogitization and exhumation of the xenoliths.