Harzburgite melting with and without H2O: Experimental data and predictive modeling
- 3 February 2004
- journal article
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research
- Vol. 109 (B2)
Abstract
The effect of H2O on harzburgite‐saturated melts has been quantified with a series of hydrous and anhydrous melting experiments using a piston‐cylinder device. Experimental conditions were 1.2–2.2 GPa and 1175–1500°C. Melt H2O contents range from 0 to 10 wt %. The effects of temperature, pressure, and bulk composition (including H2O) on the SiO2 content of the experimental melts have been evaluated using SiO2 activity coefficients. The results suggest a two‐lattice‐type model for the melt phase in which H2O mixes nearly ideally with other network modifiers (MgO, FeO, etc.) but does not mix on the network‐forming lattice site and so has little effect on SiO2 activity coefficients. The effect of H2O on SiO2 activity is too small to produce the high SiO2 contents observed in mafic andesite magmas. It is proposed that the SiO2‐rich character of hydrous, subduction‐related magmas is the result of the low temperatures at which hydrous melting occurs relative to anhydrous melting. Partition coefficients for MgO and FeO increase at lower temperatures, while the partition coefficient for SiO2 is nearly constant and is buffered by olivine‐orthopyroxene equilibria. Therefore the SiO2/(MgO + FeO) ratios of harzburgite saturated melts increase as temperature falls in both hydrous and anhydrous systems. The results suggest that H2O contents of andesitic magmas may be far higher (>7 wt %) than is generally accepted. Experimentally measured mineral/melt partition coefficients (this study and literature data) have been parameterized in terms of pressure, temperature, and melt H2O content. These expressions have been used to construct a Gibbs‐Duhem‐based numerical model that predicts the compositions of hydrous and anhydrous olivine‐orthopyroxene‐saturated melts. Comparisons with experimental data not included in the model indicate that it is the most accurate model available for predicting the compositions of high‐degree mantle melts, with or without H2O.Keywords
This publication has 61 references indexed in Scilit:
- The andesite aqueduct: perspectives on the evolution of intermediate magmatism in west-central (105–99°W) MexicoContributions to Mineralogy and Petrology, 2002
- Melting relations of the hydrous primitive mantle in the CMAS–H2O system at high pressures and temperatures, and implications for generation of komatiitesPhysics of the Earth and Planetary Interiors, 2001
- Solubilities of noble metals in Fe-containing silicate melts as derived from experiments in Fe-free systemsAmerican Mineralogist: Journal of Earth and Planetary Materials, 2000
- Algorithmic modifications extending MELTS to calculate subsolidus phase relationsAmerican Mineralogist: Journal of Earth and Planetary Materials, 1998
- An analysis of variations in isentropic melt productivityPhilosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 1997
- Compositions of near-solidus peridotite melts from experiments and thermodynamic calculationsNature, 1995
- Determining the composition of high-pressure mantle melts using diamond aggregatesGeochimica et Cosmochimica Acta, 1994
- Primitive basalts and andesites from the Mt. Shasta region, N. California: products of varying melt fraction and water contentContributions to Mineralogy and Petrology, 1994
- Water and magmas; a mixing modelGeochimica et Cosmochimica Acta, 1975
- Evidence for a Picritic, Volatile-rich Magma beneath Mt. Shasta, CaliforniaJournal of Petrology, 1974