An experimental study of the trace element geochemistry of zircon, and its use as a proxy for the oxidation state and crystallisation temperature of crustal magmas

Abstract

Excursions in Ce and Eu concentrations in zircon (ZrSiO4) relative to the smooth trend with ionic radius defined by the concentrations of the other rare earth elements (REE) arise due to the variable oxidation states of Ce (3+ and 4+) and Eu (2+ and 3+) relative to the exclusively trivalent state of the other REE. The magnitude of these anomalies is primarily a function of oxygen fugacity (fO2), so zircon has powerful potential as an indicator of magma redox state. However, the effect of temperature, pressure and the composition of the melt on Ce and Eu anomalies in zircon are largely unknown and have limited the applicability of this tool to natural samples. To examine the major controls on REE partitioning in zircon and to interpret differences in the magnitude of Ce and Eu anomalies, zircon was grown experimentally from a series of silicate melts at controlled temperature, pressure and fO2. The concentrations of the REE in zircon crystals, which were less than the minimum LA-ICP-MS analysis spot, were determined using a two-component regression method. The resulting dataset, which includes the results of over 350 zircon-melt partitioning experiments, reveals that REE uptake in zircon is favoured at low temperatures and from more evolved melts, and with increasing fO2 Ce becomes more, and Eu less, compatible in zircon. An expression was produced that relates the Ce anomaly in zircon to all the experimental variables and allows fO2 to be predicted within two log units. This is an improved geochemical tool for estimating the redox state of intermediate to felsic melts, has value as a discriminator of porphyry-Cu ore-bearing potential, and can provide insights into magmatic processes that occurred in the early Earth. The results from these zircon-melt partitioning experiments also allow compositional heterogeneity between coeval growth sectors of zircon to be quantified. The zircon-melt partitioning of the (REE+Y)3+ is strongly dependent on the zircon growth sector, whereas there is almost no effect on the partitioning of Ce4+. The apparent Ce anomaly will thus vary with the part of the crystal analysed, and this is the major limitation of zircon as an oxybarometer. The partitioning experiments also allowed an improved model for zircon saturation in silicate melts to be developed. This model includes the data from this experimental study plus 400 data points from the literature, so is now applicable to a wider range of conditions, and reproduces zircon saturation values within 4 %, compared to 20 % for the previous models. The Ti content of zircon can be used as a geothermometer and zircon co-existing with rutile and silicate melt was grown at 800 to 1500 o C and 1 atm to 6.0 GPa to quantify the effect of temperature and pressure on Ti solubility in zircon, as well as Zr solubility in rutile. In addition to the primary effect of temperature, Ti solubility decreases by one order of magnitude over the pressure range investigated. A new model for the Ti-in-zircon thermometer that includes the effect of pressure was developed, and predicts zircon crystallisation temperatures within ± 35 o C, compared to ± 175 o C when using the previous model. Ti site occupancy in zircon is important for constraining the Ti-in-zircon geothermometer and could change from the Si-site to the Zrsite with increasing pressure. The Ti site was determined by Ti K-edge XANES spectroscopy for zircons grown from a flux at 1400 o C and 1 atm to 6.5 GPa. No evidence for a site change was observed at crustal pressures but a change does occur above 4.5 GPa. Together, the results from this thesis provide a significant contribution to the understanding of trace element uptake in zircon, which is essential for interpreting magmatic processes in crustal rocks. Show more