Research highlight

Probing Iron in Earth’s Core with Molecular-Spin Dynamics

A machine-learned molecular-spin dynamics model for iron treats longitudinal spin fluctuations explicitly, resolving the high-pressure BCC phase and the elastic and transport properties that bear on Earth’s geodynamo.

Ordinary molecular dynamics treats atoms as classical particles and leaves magnetic moments out, even though magnetism affects phase stability and mechanical response. The authors build a molecular-spin dynamics model for iron in which atomic spins evolve alongside the ionic degrees of freedom, with longitudinal spin fluctuations treated explicitly. The model is trained on first-principles data and aimed at the pressure and temperature conditions of Earth’s core.

The simulations recover the metastable body-centered-cubic phase of iron above 240 GPa and reproduce compressional wave velocities, bulk moduli, viscosity and electrical conductivity. By coupling the spin dynamics to time-dependent density functional theory, the authors also obtain electronic transport properties. These results help investigate the material behaviour relevant to the geodynamo and to iron phases near the melting curve.