On the basis of both natural samples and experimental studies, clinopyroxene is a potential reservoir for potassium in the Earth's mantle. The amount of K partitioning into clinopyroxene depends on the phase assemblage present, the bulk composition, pressure, and temperature. To investigate some of these dependencies, subsolidus and melting phase relations in the system phlogopite-diopside have been studied to 17 GPa. In this system, phlogopite becomes unstable with increasing pressure, breaking down to potassium richterite, which in turn breaks down to another K-bearing hydrous phase (phase X), such that a K-rich phase coexists with clinopyroxene to 17 GPa. Clinopyroxenes contain < or =1.3 wt% K 2 O in assemblages of phlogopite + clinopyroxene + or - olivine + or - liquid at 3-5 GPa, phlogopite + clinopyroxene + garnet + or - olivine + or - liquid at 7-9 GPa, clinopyroxene + garnet + olivine + or - potassium richterite + or - liquid at 11 GPa, and clinopyroxene + olivine + garnet + phase X at 14 and 17 GPa. In these assemblages, K is partitioned into hydrous phases or liquid, rather than into the clinopyroxene. By inference, phlogopite (or its higher-pressure breakdown products) is the primary host for K in the mantle (if H 2 O is present), and any coexisting clinopyroxene has very low concentrations of K. Conversely, the natural occurrence of clinopyroxene with >>1 wt% K 2 O requires that phlogopite, potassium richterite, or phase X is not stable in the local source environment of such samples.

You do not currently have access to this article.