HR: 1330h
AN: V42A-0325    [PDF]
TI: The Partial Molar Volume and Thermal Expansivity of Fe2O3 in Alkali Silicate Liquids: Evidence for the Average Coordination of Fe3+
AU: * Liu, Q
EM: qiongl@umich.edu
AF: The University of Michigan, Department of Geological Sciences, 2534 C.C.Little Building 425 E. University Ave., Ann Arbor, MI 48109-1063 United States
AU: Lange, R
EM: becky@umich.edu
AF: The University of Michigan, Department of Geological Sciences, 2534 C.C.Little Building 425 E. University Ave., Ann Arbor, MI 48109-1063 United States
AB: Ferric iron is an important component in magmatic liquids, especially in those formed at subduction zones. Although it has long been known that Fe3+ occurs in four-, five- and six-fold coordination in crystalline compounds, only recently have all three Fe3+ coordination sites been confirmed in silicate glasses utilizing XANES spectroscopy at the Fe K-edge (Farges et al., 2003). Because the density of a magmatic liquid is largely determined by the geometrical packing of its network-forming cations (e.g., Si4+, Al3+, Ti4+, and Fe3+), the capacity of Fe3+ to undergo composition-induced coordination change affects the partial molar volume of the Fe2O3 component, which must be known to calculate how the ferric-ferrous ratio in magmatic liquids changes with pressure. Previous work has shown that the partial molar volume of Fe2O3 (VFe2O3) varies between calcic vs. sodic silicate melts (Mo et al., 1982; Dingwell and Brearley, 1988; Dingwell et al., 1988). The purpose of this study is to extend the data set in order to search for systematic variations in VFe2O3 with melt composition. High temperature (867-1534$\deg$C) density measurements were performed on eleven liquids in the Na2O-Fe2O3-FeO-SiO2 (NFS) system and five liquids in the K2O-Fe2O3-FeO-SiO2 (KFS) system using Pt double-bob Archimedean method. The ferric-ferrous ratio in the sodic and potassic liquids at each temperature of density measurement were calculated from the experimentally calibrated models of Lange and Carmichael (1989) and Tangeman et al. (2001) respectively. Compositions range (in mol%) from 4-18 Fe2O3, 0-3 FeO, 12-39 Na2O, 25-37 K2O, and 43-78 SiO2. Our density data are consistent with those of Dingwell et al. (1988) on similar sodic liquids. Our results indicate that for all five KFS liquids and for eight of eleven NFS liquids, the partial molar volume of the Fe2O3 component is a constant (41.57 ñ 0.14 cm3/mol) and exhibits zero thermal expansivity (similar to that for the SiO2 component). This value was obtained in a fit to a linear volume equation in which the other oxide components have the following fitted partial molar volumes (cm3/mol) at 1100$\deg$C: SiO2 = 26.85$\pm$0.04, Na2O = 26.57$\pm$0.07, K2O = 42.34$\pm$0.10, and FeO = 12.84$\pm$0.28, and the following fitted fitted partial molar thermal expansivities (10-3 cm3/mol-K): Na2O = 7.73$\pm$0.12, K2O = 11.99$\pm$0.24, and FeO = 2.88$\pm$1.22. For the three sodic liquids not included in this regression, the most iron-rich (18.2 mol% Fe2O3) has a value for VFe2O3 of 44.1 cm3/mole, whereas the most iron-poor (4.4 mol% Fe2O3) has a value for VFe2O3 of 37.0 cm3/mole. This trend may reflect a greater proportion of four-fold ferric iron in iron-rich liquids, which mirrors the trend of increasing ferric-ferrous ratios in sodic liquids as a function of total iron content (Lange and Carmichael, 1989). The most polymerized liquid in our data set was a sodic liquid that has a value for VFe2O3 of 45.0 cm3/mole. It thus appears that most (13 of 16) of our experimental liquids, which span a wide compositional range, lead to a VFe2O3 (41.6 cm3/mol) which is constant with composition and temperature. However, there are three important outliers that may have implications for the appropriate value to apply to magmatic liquids.
DE: 1749 Volcanology, geochemistry, and petrology
DE: 3630 Experimental mineralogy and petrology
DE: 3640 Igneous petrology
DE: 5754 Physical properties of materials
DE: 6045 Physics and chemistry of materials
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting