Oceans of Water in the Earth's core
Abstract
While the debate on the chemical composition of the outer core continues unabated, it is essential that we consider water as an important core substance for the reasons that it is abundant and suitably light to decrease the density and melting temperature of iron. Furthermore, if water is indeed stored in the core, it should provide a suitable source for water in the mantle and affect the dynamics of melting everywhere. We have conducted several experiments to understand the iron-water chemistry at high pressure and temperatures. Several studies aimed at demonstrating the formation of hydrides at high pressures have clearly established that iron hydride forms stably at pressures from a few to 7.5 GPa both in the Fe-H2 and the Fe-H2O systems, which has led to the important proposal of including water (and thereby hydrogen and oxygen) in the core. The information on temperature effect on these systems is available to 1800 K at low pressures. With this work, we have extended the P and T range of the previous studies on iron-water reaction to nearly 85 GPa and 2000 K respectively. We used various types of diamond-anvil cells which include Mao-Bell cell for single side laser heating or for wire heating and Merrill-Basett design for double-side laser heating. In addition to experiments with laser heating, we have also conducted experiments in Mao-Bell cells with external heating. In all experiments reported here, in situ x-ray data were collected on heated samples at GSECARS and at ESRF. In presence of excess iron, which was the case for all experiments with direct wetting of iron with water in the externally heated cells, we invariably found the three phases, Fe, iron hydride and wuestite. In presence of brucite or water and MgO, we can demonstrate that melting of the hydride and the FeO component of wuestite occurred at a rather low temperature of 1525 K or less. The following inferences can be drawn based on the experimental results: a) iron-hydride stability extends to high pressures (~85 GPa) and temperatures (~1800 K) and therefore it could form in a primitive iron core and become a part of the melt if temperature exceeds the melting temperature (~2000 K) at the outer-core press, and b) in the system iron-water, the hydride phase cannot exist without wuestite and therefore both hydrogen and oxygen components will be part of the melt in the outer core. An additional important result is that we could not determine the influence of water on direct melting of iron because of the hydride and oxide reactions that preceded melting. Wuestite (FexO) or ferropericlase (a solid solution between periclase and wuestite) melted below 1525 K in presence of water. Wuestite is estimated to be present in abundance in the mantle and could be an important constituent of the primitive earth forming by direct reaction between iron and water. Although in this experiment, we can only ascertain that Fe was first oxidized and then melted, it does give us an estimate of the possible effect of water on the melting of Fe. According to Boehler, dry FeO melts at a temperature of ~2500 K at a pressure of 35 GPa. Thus water reduced the melting temperature of the pure phase by close to 1000 K. If the melting temperature of iron is similarly reduced in a fluid saturated system, the effect could be large; in a less saturated system it may lower the melting by a few hundred degrees. If even a percent of water in the core will amount to ten times more water than that present in the oceans.
- Publication:
-
AGU Spring Meeting Abstracts
- Pub Date:
- May 2002
- Bibcode:
- 2002AGUSM.M42A..04S
- Keywords:
-
- 3924 High-pressure behavior;
- 1015 Composition of the core;
- 1212 Earth's interior: composition and state (8105)