Some Remarks on the Theory of Element Partitioning

Abstract:

Element partitioning is probably the most important concept in geochemistry. However some fundamental issues on element partitioning are poorly understood or misinterpreted or misused. In this presentation, I will discuss two issues. First, the dependence of (trace) element partitioning on the properties of trace element has been understood incompletely (or incorrectly). A strain energy model is a good approximation, but in all previous models, only the influence of the size of trace element is considered. As a result, the influence of “stiffness” of the trace element was ignored and the physical meaning of elastic constants in the theory was vaguely described as “site-elasticity”. I developed a modified model in which the influence of the stiffness of the trace element is included with the correct boundary conditions. The incorporation of the influence of stiffness of trace element is critical particularly in a case where the trace element is soft, e.g., a case of noble gas. The new model provides better explanation of the observed element partitioning. Second, the influence of selective dissolution of a trace element in the melt is often ignored. It is well known that the solubility of noble gas in the melt strongly depends on its atomic size. In contrast, the dependence of trace element solubility on its size is not well known but usually ignored. I developed a conceptual model to explain why noble gas solubility is strongly size dependent but not for the solubility of other charged trace elements.

These two shortcomings in the previous models are serious when the partitioning of noble gas elements is considered. Some applications of these models will be discussed including the behavior of the noble gases during melting and solidification and the evaluation of water content and its influence on the rheological properties of the lithosphere.