A principle aim in geochemistry is to determine the chemical composition of the Earth's various reservoirs, (e.g. oceans, crust, mantle and core) and to measure the fluxes of chemical species between these regions. The geochemical investigation of the Earth's interior is mainly focused on determining the composition of the Earth's mantle and core, but faces the significant problem that samples exist only from the uppermost regions of the mantle; the core is entirely inaccessible to sampling. This problem can be approached by first estimating the composition of the bulk Earth from chondritic meteorites, which are considered to represent the undifferentiated material from which the Earth was built. Then, by performing experiments to understand how this bulk Earth composition may have fractionated during large scale mantle melting and core formation, we can attempt to piece together the likely compositions of the ensuing reservoirs.
Several of the studies reported in this section attest to the very slow rate of chemical diffusion in high-pressure mantle minerals. Consequently many fractionation processes are likely to take place too rapidly for chemical equilibrium to be achieved. In such cases experimental studies of reaction kinetics are required in order to estimate the compositions of forming reservoirs. An example is the separation of metallic liquid from the silicate mantle during core formation. Determining the extent to which descending diapirs of metallic liquid equilibrated with the surrounding solid lower mantle can place important constraints on the likely chemistry of the core and the mechanisms of core formation.
Another consequence of slow diffusion in silicates is that it inhibits the remixing of chemical heterogeneities in the mantle, such as those resulting from the subduction of basaltic oceanic crust. Although subducted crust is mixed back into the mantle by convection on a large scale, small regions of crust are likely to persist due to the immense time scales required for them to dissipate by diffusion. Studies to understand the likely mineralogy and chemistry of possible mantle heterogeneities are very important because such regions may have a disproportionate effect on trace element concentrations and sensitive isotopic systems such as U/Pb and Re/Os. A further possibility is that small-scale heterogeneities may even remain in the mantle as a remnant of crystal settling from a primordial magma ocean.
Although most mantle samples originate from only the top 200 km of the upper mantle, in recent years some diamonds with silicate and oxide inclusions have been proposed to originate from much greater depths in the transition zone and lower mantle. To get a clearer understanding of the expected chemistry of minerals in the lower mantle several of the studies in this section are aimed at determining the partitioning of Fe (Fe2+ and Fe3+) between lower mantle minerals. This may also help to determine the likely conditions of formation of proposed deep mantle diamond inclusions. If such inclusions really do reflect a lower mantle origin, this will significantly expand our mantle sampling depth to over 600 km.