Experimental Investigation: The Influence Of Small Amounts Of Water On Peridotite Melting In The Oceanic Upper Mantle

Glenn Gaetani , Geology & Geophysics



The nominally anhydrous minerals that comprise mantle peridotite - olivine, high-Ca clinopyroxene, and orthopyroxene - can accommodate enough water into their structures to strongly influence partial melting of the mantle beneath mid-ocean ridges. At a given mantle temperature, the presence of only ~50 to 200 ppm water dissolved in the oceanic upper mantle causes magma generation to begin at significantly greater depths than would be the case for truly anhydrous peridotite, leading to an increase in the amount of melt produced and increased thickness of the resulting oceanic crust. Therefore, understanding the process of melt generation beneath mid-ocean ridges and quantifying relationships among mantle temperature, extent of peridotite partial melting, and the thickness of the oceanic crust requires a thorough understanding of the influence of small amounts of water on the melting point (solidus) of peridotite. However, the depth at which melting begins for a given mantle temperature and water content are poorly constrained because the influence of small amounts of water on the peridotite solidus has never been experimentally determined. This is attributable largely to the technical difficulties associated with (1) precisely controlling small concentrations of water in highpressure melting experiments, and (2) detecting the presence of small amounts of partial melt in an experimental charge. Here I propose a new approach to experimentally determining the effect of small amounts of water on the peridotite solidus that is designed to overcome these hurdles. The major difficulty associated with precisely controlling low concentrations of water in highpressure experiments is that the powdered starting material is capable of adsorbing more water than is required for the experiment. Adsorbed water can be driven off by heating the powder at 400 °C for 12 hours and immediately welding the sample capsule shut. I propose to synthesize a peridotite from high-purity reagents and add a controlled amount of water as talc, a mineral that is stable at high temperatures and contains ~5 wt% structurally bound water. Experimental charges can then be heated at 400 °C to drive off the adsorbed water without losing the structurally bound water needed for the experiment. Detecting the presence of small amounts of partial melt in the experiments will be accomplished by using secondary ion mass spectrometry to determine the amount of water dissolved in the peridotite minerals at a given pressure and temperature. At conditions below the peridotite solidus, all of the water will be hosted in the minerals. At conditions above the peridotite solidus, water will partition strongly into the melt and the concentration in the minerals will drop considerably. By carrying out experiments over a range of temperatures at a given pressure and determining the water content of the minerals, the melting point of the peridotite can be determined to within ~10 °C. Funds from this proposal will be used to carry out a pilot study to demonstrate that this approach works, and the data will form the basis for a full proposal to NSF Ocean Sciences. Given that the experimental community is convinced that these difficulties cannot be overcome, a pilot study is a necessary prerequisite for successfully obtaining NSF funding