HR: 16:00h
AN: V52F-01 [PDF]
TI: Experimental Investigation of the Partially-Hydrated Solidus of KLB-1
AU: * Baker, L J
EM: labaker@gps.caltech.edu
AF: California Institute of Technology, M/C 170-25, Pasadena, CA 91125 United States
AU: Asimow, P D
EM: asimow@gps.caltech.edu
AF: California Institute of Technology, M/C 170-25, Pasadena, CA 91125 United States
AB:
It is important to study the effects of the presence of trace amounts of water, found in nominally anhydrous minerals, on
mantle melting by more accurately determining the position of the peridotite solidus as a function of trace water content at
upper mantle pressure and temperature conditions.
The specific effects of trace water on melt productivity and the composition of liquids in equilibrium with mantle minerals
can be examined using experimental methods. However, there is a lack of experimental data with regard to the trace water
contents associated with the mid-ocean ridge basalt source mantle (50 to 500 parts per million H$_{2}$O) as opposed to fully
saturated or nominally anhydrous conditions. The complex speciation of H$_{2}$O dissolution in silicate liquids makes it
difficult to extrapolate from either dry or saturated runs into this slightly hydrous regime. The challenges to accurate
experimental measurement of the "partially hydrous" peridotite solidus include control of initial and final water contents
and characterization of extremely small melt fractions.
An experimental protocol has been developed to minimize external water contamination. This includes a rigorous drying
sequence to eliminate adsorbed water from the system. Our double capsule assembly utilizes a hematite "hydrogen-getter" in
order to discourage hydrogen diffusion into the sample during the run. Time-series experiments using an anhydrous glass of
albite-orthoclase eutectic composition constrain the amount of water entering the system from the environment. We found that
no more than 280 ppm can contaminate the system during run durations of up to 72 hours using this assembly at P = 1.9 GPa
and T = $1350\deg$C.
Peridotite melting experiments were performed using the piston-cylinder device at 2.0 GPa. The sample material is a mix
compositionally equivalent to the fertile spinel peridotite KLB-1 comprised of a synthetic component commensurate to the
pyroxene plus spinel portion in addition to natural KLB-1 olivine material. The olivine component was doped with H$_{2}$O in
water-saturation experiments, allowing the accurate representation of peridotite composition as well as allowing for a trace
amount of H$_{2}$O. The incorporation of water as measurable structurally bound hydroxyl within the olivine enables the
aggressive drying protocol to remove additional adsorbed water while still permitting a pre-determined amount of water to
remain in the system.
In order to determine precisely the position of the solidus and to evaluate near-solidus melt productivity, a method is
required by which the presence of a very small degree of melt ($<$1%) can be recognized. Simple observation of the presence
or absence of glass or quench material may not be sufficient at such low melt fractions. To this end, analyses of the
incompatible elements Na and Ti in clinopyroxene grains have been performed using a trace element protocol on the EPMA. The
onset of melting is demonstrated by the sudden drop in concentration of the element in the clinopyroxene, allowing a more
accurate estimation of the position of the solidus than had been previously measured. A preliminary estimate for the
position of the anhydrous solidus of KLB-1 at 2.0 GPa is between 1325-$1350\deg$C. Applying a similar method with a more
incompatible tracer (e.g. H or K) is expected to reveal the slightly hydrated solidus.
DE: 1025 Composition of the mantle
DE: 1094 Instruments and techniques
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting