17:15 〜 19:15
[SIT20-P14] Water partitioning between wadsleyite and garnet under water-unsaturated conditions
キーワード：水分配、ウォズレアイト、ガーネット、水雰囲気
Nominally anhydrous minerals contain small amounts of structurally incorporated water in the form of hydroxyl. This water significantly influences their physical and chemical properties despite its low concentration. Previous high-pressure, high-temperature experiments have determined the water solubility and partition coefficients of these minerals under water-saturated conditions. However, water partitioning may vary with water fugacity, particularly for minerals with multiple water incorporation mechanisms. As the mantle is unlikely to be fully water-saturated, conventionally used partition coefficients may not accurately reflect real mantle conditions. To better understand water partitioning in the upper mantle, it is crucial to determine water partition coefficients under water-undersaturated conditions.
In this study, we investigated the water partitioning coefficient between wadsleyite (Wds) and garnet (Grt) under water-undersaturated conditions while minimizing grain boundary effects. To achieve this, large single crystals of both minerals were synthesized in a single run using a K2CO3 flux. The K2CO3 component diluted water in the hydrous melt to reduce the water fugacity in the system. Although obtaining large crystals under low water-fugacity conditions is challenging, the K2CO3 flux promoted the growth of large single crystals, which allowed us to measure the water content in these minerals using FT-IR spectroscopy without grain-boundary effects.
The crystal growth experiment was performed using a Kawai-type multi-anvil apparatus at the Bayerisches Geointitut (BGI), University of Bayreuth. The starting material for Wds and Grt was a mixture of silicates. The chemical composition of the silicate source was prepared by adjusting a mixture of powders to achieve a 1:1 molar ratio of forsterite to enstatite in an iron-free system, with an appropriate amount of Al2O3 added. The ratio of the starting material to the K2CO3 flux is 15:1 to 6:1 by volume. The flux was not particularly doped with water, as it contained some water from the air during the preparation of the experiment. The run conditions were a pressure of 16 GPa, a temperature of 1873K, and a duration of a maximum of 5 hours.
We have succeeded in simultaneously synthesizing single crystals of both Wds and Grt in water-rich and water-poor runs. The grain sizes of Wds and Grt were 300 and 500 mm, respectively, in both water-rich and –poor run. The phase was identified and the crystals were oriented using a single-crystal X-ray diffractometer at BGI. Their water contents were measured on the oriented sections of Wds and Grt using a polarized Fourier transform infrared spectroscopy (FT-IR) at the Geodynamic Research Center (GRC), Ehime University. The partition coefficient of water between Wds and Grt and, was DH2OWds-Grt ≈ 10 in the water-rich run and DH2OWds-Grt ≈ 0.8 in the water-poor run, respectively. Thus, we have first demonstrated that the water partitioning between Wds and Grt varies with water fugacity. It should be noted that the number of runs is only two, and therefore, more data points are desirable to reach a robust conclusion.
The current results suggest that although water was previously thought to be partitioned selectively into Wds and affect its properties, water is partitioned almost equally into both minerals and affects the properties of both minerals in the real mantle. For example, the conventional view is that water should soften the harzburgite layer in subducted slabs, but leave the basaltic layer hard, whereas the present results suggest that water should soften both layers.
In this study, we investigated the water partitioning coefficient between wadsleyite (Wds) and garnet (Grt) under water-undersaturated conditions while minimizing grain boundary effects. To achieve this, large single crystals of both minerals were synthesized in a single run using a K2CO3 flux. The K2CO3 component diluted water in the hydrous melt to reduce the water fugacity in the system. Although obtaining large crystals under low water-fugacity conditions is challenging, the K2CO3 flux promoted the growth of large single crystals, which allowed us to measure the water content in these minerals using FT-IR spectroscopy without grain-boundary effects.
The crystal growth experiment was performed using a Kawai-type multi-anvil apparatus at the Bayerisches Geointitut (BGI), University of Bayreuth. The starting material for Wds and Grt was a mixture of silicates. The chemical composition of the silicate source was prepared by adjusting a mixture of powders to achieve a 1:1 molar ratio of forsterite to enstatite in an iron-free system, with an appropriate amount of Al2O3 added. The ratio of the starting material to the K2CO3 flux is 15:1 to 6:1 by volume. The flux was not particularly doped with water, as it contained some water from the air during the preparation of the experiment. The run conditions were a pressure of 16 GPa, a temperature of 1873K, and a duration of a maximum of 5 hours.
We have succeeded in simultaneously synthesizing single crystals of both Wds and Grt in water-rich and water-poor runs. The grain sizes of Wds and Grt were 300 and 500 mm, respectively, in both water-rich and –poor run. The phase was identified and the crystals were oriented using a single-crystal X-ray diffractometer at BGI. Their water contents were measured on the oriented sections of Wds and Grt using a polarized Fourier transform infrared spectroscopy (FT-IR) at the Geodynamic Research Center (GRC), Ehime University. The partition coefficient of water between Wds and Grt and, was DH2OWds-Grt ≈ 10 in the water-rich run and DH2OWds-Grt ≈ 0.8 in the water-poor run, respectively. Thus, we have first demonstrated that the water partitioning between Wds and Grt varies with water fugacity. It should be noted that the number of runs is only two, and therefore, more data points are desirable to reach a robust conclusion.
The current results suggest that although water was previously thought to be partitioned selectively into Wds and affect its properties, water is partitioned almost equally into both minerals and affects the properties of both minerals in the real mantle. For example, the conventional view is that water should soften the harzburgite layer in subducted slabs, but leave the basaltic layer hard, whereas the present results suggest that water should soften both layers.