Partitioning between mantle crystalline phases and melt has a major control on argon systematics and the evolution of the planetary atmosphere and large scale differentiation of the earth. Experimental partitioning data can potentially provide insights into differentiation and degassing history of magmas. There has been a great deal of controversy regarding these coefficients, with values ranging from
3-4 orders of magnitude. A new technique, the UVLAMP (Ultra-Violet Laser Ablation Micro-Probe), avoids problems commonly encountered using bulk extraction techniques, and offers the chance to obtain meaningful diffusion data. As olivine and clinopyroxene play a dominant role in degassing processes, our initial studies have focused on the olivine-melt and clinopyroxene-melt partitioning of argon.
To this end, olivine and clinopyroxene crystals were grown by slowly cooling synthetic basalt melts (from 1320°C and 1285°C) in a furnace at one atmosphere of argon pressure. A UVLAMP was used to ablate small regions of pure olivine, clinopyroxene, and melt, and to measure the quantities of 40Ar and 36Ar released. These quantities are used to calculate crystal/melt partition coefficient values for argon in experimental basalt systems. Where possible, the UV laser was rastered over square regions (up to 150 µm2 x 95 µm deep) ablating 10-15 µm per layer resulting in depth profiles. This allowed removal of the surface absorbed argon effect.
In addition to the olivine-melt and clinopyroxene-melt experiments, a natural haplobasaltic glass has been synthesised at 1 kbar of argon pressure and varying temperatures (up to 675°C). Two depth profiling techniques were employed to determine diffusion rates in the glasses. The measured diffusivity of argon in the haplobasaltic glass is similar to values quoted for silica and orthoclase composition glasses (Carroll and Stolper, 1991; Carroll, 1991).
One important result of these experiments is the determination of a clinopyroxene/melt partition coefficient for argon of 0.0008 ± 0.0005. This value is three orders of magnitude lower than previous experiments using the bulk analysis technique carried out on the same sample (i.e. >0.15) (Broadhurst et al, 1990, 1992).
The results show that the UVLAMP technique is an important tool for determining solubilities and partition coefficients of noble gases in silicate melts and crystals. This 'in situ' micro-technique has several advantages over the common bulk analytical methods: (a) no need to separate crystals and melt, (b) precise location of small volumes of sample, avoiding gas-rich inclusions, and, (c) removing surface absorbed argon effects.
Broadhurst, C.L., Drake, M.J., Hagge, B.E. & Bernotowicz, T.J., Geochim. Cosmochim. Acta 54, 299-309 (1990).
Broadhurst, C.L., Drake, M.J., Hagge, B.E. & Bernotowicz, T.J., Geochim. Cosmochim. Acta 56, 709-723 (1992).
Carroll, M.R., Earth Planet. Sci. Lett. 103, 156-168 (1991).
Carroll, M.R. & Stolper, E.M., Geochim. Cosmochim. Acta 55, 211-225 (1991).