Boron partitioning between zircon and melt: Insights into Hadean, modern arc, and pegmatitic settings

Abstract

There is a lot of debate surrounding the geodynamic environment of the Hadean Earth, and the only definitively known fragments from the first 500 Ma are detrital zircons. This makes the combination of natural and experimentally obtained data a powerful tool to unravel the chemistry of the early Earth. We have used an incompatible marker element, B, to partially derive the chemistry of the parent melts of the Hadean and modern zircons. We report an experimental calibration for temperature dependent B partitioning between zircon and a hydrous weakly peraluminous granitic melt.log10DBzrc/melt=−1027±372TK−2.011±0.257where DBzrc/melt is the zircon-melt partition coefficient for B and T is temperature in K.This calibration has been applied to natural samples, viz., Hadean zircons from the Jack Hills (JH), Australia, Phanerozoic zircons from the Lachlan Fold Belt (LFB), Australia and three pegmatitic zircons from Seiland Igneous province, Norway, Paicoma Canyon, California and Freeman Mine, North Carolina. Our results present direct evidence of B being present in the Hadean crust. The zircons from JH and LFB are rather poor in B (8–80 ppb), but comparable to each other, while the pegmatites have as much as ten times the [B] (~0.35–0.45 ppm). Application of our experimental calibration yields calculated B melt concentrations of 10–90 ppm for the JH and LFB zircons. Such values for melt [B] (concentrations) are similar to the modern upper continental crust (17 ppm) and volcanic arcs, but are high when to compared to OIBs (0.6–1.8 ppm) and MORBs (1.3 ppm). The Lachlan zircons have [B] values that are broadly similar to the detrital Hadean and Archean zircons, and these have been presented as a point of comparison between Hadean and modern zircons. The pegmatite zircons return calculated melt values (244–701 ppm) that are much higher than the parent melts of the Australian zircons. One of the pegmatite zircons show a variation in calculated Ti-in-zircon crystallization T from the core to the rim (716 °C – 916 °C). Two of the other zircons crystallized at 664 ± 14 °C and 598 ± 13 °C (2 s.e.; not taking into consideration a ~50oT uncertainty due to imperfectly constrained silica/titania activities) and show no intracrystalline variation in [B] or T. Finally, B has been proposed to have been a possible stabilizing agent for ribose aqueous solutions and could have played a part in the formation of carbohydrates and proteins which were building blocks for RNA. Our documented presence of B in the Hadean crust at calculated concentrations similar to a modern volcanic arc setting, makes the role of B in ribose stabilization at least possible on the primordial Earth.