Hydrogen concentration in plagioclaes as a hygrometer of magmas: Approaches from melt inclusion analyses and hydrous melting experiments
Abstract
Plagioclase is one of the nominally anhydrous minerals (NAMs) which accommodates hundreds wt. ppm of hydrogen. Hydrogen in igneous plagioclase (OH) can act as a proxy of dissolved H2O in silicate melt. In order to use it as a practical hygrometer of magmas, we studied partitioning of hydrogen between plagioclase and basaltic melt dis-solving H2O (0.3∼5.5 wt.%) by two approaches: analyses of plagioclase-hosted melt inclusions (H2O≈0.3 wt.%) from mid-ocean ridge basalt (MORB) and hydrogen partitioning experiments between An96 plagioclase and hydrous basaltic melt (0.8 wt.%≦H2O≦5.5 wt.%) at 0.35 GPa. Concentration hydrogen in plagioclase and concentration of H2O in basaltic glasses were analyzed by infrared spectroscopy. As a first series of this study, plagioclase-hosted melt inclusions in MORB (50.5% SiO2, 15.1% Al2O3, 7.4% MgO) from the Rodriguez triple junction in the Indian Ocean were analyzed. The hydrogen concentration of plagioclase is less than 50 wt. ppm water, and no correlation between hydrogen concentration and anorthite content is recognized. Average H2O concentrations in melt inclusions is 0.3 wt.%. Therefore, apparent partition coefficient of hydrogen between plagioclase and melt is ≈ 0.01 in molar basis. As a second series of this study, hydrous melting experiments of arc basaltic magma were performed at 350 MPa using internally-heated pressure vessel. Starting material was hydrous glass (0.8 wt.%≦H2O≦5.5 wt.%) of an undifferentiated rock from Miyakejima volcano, a frontal-arc volcano in Izu-arc (MTL rock: 50.5% SiO2, 18.1% Al2O3, 4.9% MgO). A grain of Ca-rich plagioclase (≈1 mg, about An96) and 10 mg of powdered glasses were sealed in Au80Pd20 alloy capsule, and then kept at temperature at around the liquidus. Liquidus phase of MTL rock at 350 MPa is always plagioclase with 0 to 5.5 wt.% H2O in melt, and therefore, a grain of plagioclase and hydrous melt were nearly in equilibrium. Oxygen fugacity (fO2) during the melting experiments was not controlled, which was estimated to be 3 log unit above Ni-NiO buffer. Experiments were quenched after 24 hours, long enough to attain equi-librium partitioning of hydrogen between plagioclase and melt. Combining two series of study, we formulated empirical equation to correlate hydrogen concentration in plagioclase and H2O concentration in melt as: Hydrogen in plagioclase (wt. ppm water) ≈ 94.3×√(H2O in melt, wt.%). We also determined partition coefficient of hydrogen between plagioclase and basaltic melt to be 0.01±0.005 under H2O-poor conditions (≤150 wt. ppm water in plagioclase, ≤2 wt.% H2O in melt) and to be 0.008±0.002 under H2O-rich conditions (≥150 wt. ppm water in plagioclase, ≥2 wt.% H2O in melt). Slight decrease in hydrogen partitioning with increasing H2O can be related to change in hydrogen site in crystal structure of plagioclase (possibly tetrahedral sites). Peak position of infrared absorbance spectra lies at 3200-3400 cm-1 under H2O-poor condition and at 3600 cm-1 under H2O-rich condition, suggesting expansion of O-H…O bond length with increasing H2O.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.V11C2776H
- Keywords:
-
- 8413 VOLCANOLOGY / Subduction zone processes;
- 8416 VOLCANOLOGY / Mid-oceanic ridge processes;
- 8430 VOLCANOLOGY / Volcanic gases