Melt Partitioning

The crystal-melt partition coefficient KD = CJCh where Cs is concentration in a solid and Q is concentration in coexisting liquid, is a key parameter in trace element studies of igneous systems. A noble gas crystal-melt partition coefficient is the ratio of the gas solubilities considered here. As seen in Table 2.3, solubilities have now been reported for a variety of melt compositions, but solubility data are still very scarce for solids in general. 

There are a number of trace elements commonly described as incompatible because their partition coefficients (for typical mantle mineralogy) are low, often as low as 10 2 or 10 or less. Prominent examples are K, Rb, U, Th, and the rare earth metals. The principal reason for their incompatibility is evidently their large ionic radii. Concentrations of these elements in mantle-derived rocks are relatively low, and the generalization emerges that they have been expelled from the mantle (at least the upper mantle) and concentrated in the crust, especially continental crust. 

It is generally believed that the noble gases should behave like incompatible elements, only more so. Their atomic radii are also relatively large (Table 2.1). More- 

Hiyagon and Ozima (1986) employed a laboratory approach of measuring crystal-melt partition coefficients. They measured noble gas concentration in olivine crystals and basalt melts, which were synthesized at 1370-1300°C under an atmospheric pressure, and also at 1360-1050°C under high pressure (0.2-1.5 GPa), of noble gas mixture. From these experimental results, they obtained ranges for noble gas partition coefficients XHe = 0.07, XNe = 0.006-0.08, KM = 0.05-0.15, KXe = 0.3. These partition coefficients are much larger than the values obtained by Marty and Lussiez (1993) and also these of common incompatible elements such as U (-0.002) or K (0.0002 - 0.008) between olivine and basalt melt (e.g. Henderson, 1982). 

A common difficulty in these partition experiments, with the use of either natural or synthesized samples, is in achieving perfect separation of the melt from the crystal phase for determining the noble gas content. Even a very small amount of glass (melt) contamination in crystal would increase the partition coefficient considerably, since noble gasses are much more enriched in glass. To circumvent this difficulty, Broad-hurst et al. (1990, 1992) prepared natural minerals and synthetic silicate melts that 

---

Mineral-Melt Partitioning of Uranium, Thorium and Their Daughters... 

Mineral-Melt Partitioning ofU-Series Nuclides Clinopyroxene along mantle solidus... 

Figure 4. Fits of lattice strain model to experimental mineral-melt partition coefficients for (a) plagioclase (run 90-6 of Blundy and Wood 1994) and (b) elinopyroxene (ran DC23 of Blundy and Dalton 2000). Different valence cations, entering the large cation site of each mineral, are denoted by different symbols. The curves are non-linear least squares fits of Equation (1) to the data for each valence. Errors bars, when larger than symbol, are 1 s.d. Ionic radii in Vlll-fold coordination are taken from Shannon (1976).

Before discussing mineral-melt partition coefficients in detail, it is useful to consider other factors that may influence partition coefficients for the U-series elements. Such factors arise both because the U-series elements typically occur at very low abundances in nature, and because they are radioactive. The first feature introduces the possibility of deviations from Henry s Law, at very low concentrations. The second feature raises... 

Orthopyroxene has a Vl-fold Ml site and a Vl-fold M2 site. Both are predominantly filled by Mg and Fe. The smaller Ml site shares many characteristics with the clinopyroxene Ml site. It is therefore reasonable to assume that no U-series cations, including Pa (see above) enter that site. We will confine our discussion to the octahedral M2, which is smaller than the equivalent (Vni-fold) clinopyroxene site, even after allowing for the different coordination number. Consequently most of the U-series elements have very low orthopyroxene-melt partition coefficients. 

By analogy with clinopyroxene it is likely that Pa enters the orthopyroxene M2 site. In light of the fact that Du and Dxh in orthopyroxene are approximately ten times lower than in clinopyroxene, it is likely that Z)pa is also lower in orthopyroxene. However, this effect is offset to some extent by the smaller M2 site in orthopyroxene, which will tend to be more favourable to Pa than the M2 site in clinopyroxene. We have used the electrostatic model, applied to the two orthopyroxene-melt partitioning experiments of McDade et al. (2003a,b) to derive (Fig. 13). Both datasets, at 1.5 and 3 GPa, are... 

Van Westrenen et al. (2001a) present a model of lanthanide and Sc partitioning between the garnet X-site and melt. The model is a variant of the lattice strain model of clinopyroxene-melt partitioning of Wood and Blundy (1997), and is based on 160 experimental garnet-melt pairs in the pressure-temperature range 2.5-7.5 GPa and 1450-1930°C. The model includes composition-sensitive expressions for and accounts for the non-linear variation in with composition, as follows ... 

---