HR: 0800h
AN: V31D-0679    [Abstracts]
TI: Vanadium Partitioning and Mantle Oxidation State: New Experimental Data
AU: * Mallmann, G
EM: guilherme.mallmann@anu.edu.au
AF: Research School of Earth Sciences, Australian National University, Building 61, Mills Road, Canberra, ACT 0200, Australia
AU: O'Neill, H S
EM: hugh.oneill@anu.edu.au
AF: Research School of Earth Sciences, Australian National University, Building 61, Mills Road, Canberra, ACT 0200, Australia
AB: Vanadium exists in multiple valences in natural basaltic melts, namely V²⁺, V³⁺, V⁴⁺ and V⁵⁺. Because most crystalline phases prefer to incorporate V³⁺ rather than V⁴⁺ and V⁵⁺, the crystal/silicate-melt partitioning of vanadium (D_V^cry/melt) tends to decrease with increasing oxygen fugacity (fO₂). Such dependence has been experimentally demonstrated and used to estimate the fO₂ of mantle and mantle-derived rocks. Recent modelling of V and V/Sc systematics in basalts has lead to the view that the relative fO₂ of the upper mantle is constant, both through time and among the sources of different types of basaltic magmas (i.e. MORB, OIB and IAB). This is contrary to the notion given by other oxygen barometric methods on peridotites and basalts, which indicate an upper mantle heterogeneous in relative fO₂. To explore further the potential of V abundances and V/Sc ratios to estimate the relative fO₂ of mantle peridotites and basalts, and in particular to understand variations in mantle oxidation state better, we carried out an experimental campaign aimed at measuring D_V^cry/melt for all the relevant phases of the upper mantle (i.e. olivine, orthopyroxene, clinopyroxene, garnet and spinel) over a range of fO₂ conditions large enough to pin down not only the behaviour of V³⁺ and V⁴⁺ but also V²⁺ and V⁵⁺. Experiments were done in 1-atm vertical tube furnaces (1300°C) and piston-cylinder apparatus (1275-1450°C and 1.5-3.2 GPa). For the high-pressure experiments, fO₂ was controlled by the Re-ReO_x/2 equilibrium (10^-9 to 10^-0.7 bar), whereas for the 1-atm experiments, fO₂ was controlled by Ar-CO-CO₂- O₂ gas mixes (10^-18 to 10^-0.7 bar). Five starting compositions were used to ensure the presence of all the desired phases. Experimental products were analysed for major elements by electron microprobe and for trace elements by laser-ablation ICP-MS, which enables V to be measured precisely even at very low concentrations. Partition coefficients for all phases plot as approximately sigmoid-shaped curves in log D-log fO₂ space. Details of the shape of the curve are controlled by the relative preference of each crystalline phase for a specific valence of V. For instance, orthopyroxene appears to particularly like V⁴⁺, so that the log D_V^opx/melt-log fO₂ and log D_V^cpx/melt-log fO₂ curves converge in the region of the diagram dominated by V⁴⁺, diverging in the regions dominated by V³⁺ and V⁵⁺. Contrary to previous studies, our results do not suggest a systematic increase in D_V^cpx/opx with decreasing fO₂. Olivine and spinel, on the other hand, strongly prefer V³⁺ relative to V⁴⁺ and V⁵⁺ and hence for olivine and spinel the difference in partition coefficients between reducing and oxidizing conditions are more pronounced than that for pyroxenes. At high-pressure, D_V^grt/melt and D_V^cpx/melt are very similar to each other, but the values of D_V^cpx/melt are about one order of magnitude higher than those obtained at 1 atm at comparable fO₂. The cause of this discrepancy is being investigated.
DE: 1009 Geochemical modeling (3610, 8410)
DE: 1025 Composition of the mantle
DE: 1037 Magma genesis and partial melting (3619)
DE: 1038 Mantle processes (3621)
DE: 1065 Major and trace element geochemistry
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2007 Fall Meeting