Effects of Fe on melts

In addition to modelling crystalline Earth materials, we will study the effects of electronic transitions on iron in silicate melts. This requires doing computations on snapshots obtained from first-principles molecular dynamics (FPMD) based on DFT or DFT+U methods. This has already been shown tractable so we can even use snapshots from trajectories already produced and as correlated methods scale linearly with correlated ions per supercell, these calculations will also be tractable on large national parallel computers.

Questions we will address are:

• How does element partitioning change between solids and melts, and under what conditions do melts segregate into coexisting iron-rich metallic melts and iron-poor insulating melts?
• Will iron rich silicate melts be metallic in the deep Earth, thus greatly changing their thermal conductivity and reactivity?
• What is the density of iron rich melts?
• How do these processes affect a magma ocean, as likely existed in early Earth history?

In addition to melts in the mantle, we will investigate the behaviour of iron-rich melts in the outer core. Many pioneering studies have been carried out using first-principles molecular dynamics, and DFT may be sufficient for studying iron-rich melts under core conditions. Still, many-body techniques will help in computing properties such as electrical conductivity and magnetic moments etc. important for modelling the geodynamo.