3/13/2012 - Zircon Saturation Re-Revisited


12:00 PM - 1:00 PM
4460 Geology

Presented By:
Patrick Boehnke - UCLA


Zircon saturation in silicate melts has been utilized for magma thermometry and predicting the survival of zircon xenocrysts in crustal melts for nearly 30 years. The original calibration, which assumed only compositional (M = [2Ca+Na+K]/[AlxSi]) and temperature controls, was bolstered by subsequent experimental investigations and thermometry of volcanic rocks and glasses. These latter studies, while confirming the general predictions of the model, suggested that other environmental parameters (e.g., pressure, H2O, halogens, [Fe], oxygen fugacity, etc.) might have second-order effects. Given the tremendous advances in micro-analytical capabilities over the intervening three decades, we have returned to this question with a view to obtaining a refined zircon solubility calibration as a function of P, T, [H2O] and FM (= [Na+K+2(Ca+Mg+Fe)]/[AlxSi]). Detailed SEM imaging of the original low-temperature crystallization experiments (1.2-2.1 kbar) revealed limitations of this approach and we chose instead to use a new experimental design in which shattered Mud Tank zircon is infiltrated by melts of selected composition and water contents. 10 kbar hydrothermal experiments (925º and 850ºC) were run for sufficiently long durations (2 to 3 days) to ensure microscale diffusive equilibration of Zr released by zircon dissolution into the intercrystalline melt pools. Sectioned run products were analyzed by SIMS ion imaging of selected areas where glass is exposed in close proximity to or surrounded by Mud Tank zircon fragments. Ion imaging has the advantage of permitting high spatial resolution (3 µm) analysis of the glasses allowing assessment of Zr equilibration. Using synthetic glass standards, we found [Zr] in anhydrous glasses to be enhanced by ca. 20% relative to hydrous (at 6 wt.% H2O). Our new experiments and re-analysis of the earlier glasses broadly reproduce the original calibration, albeit with substantially enhanced (factor of five) precision compared to the original EMPA analyses. Thus it appears that no significant pressure effect exists up to at least 10 kbar. Ongoing work will expand the pressure range beyond this limit and explore a greater compositional space than previously constrained.

Return to Seminar Schedule