Is the inner core of the Earth pure iron?

The properties of the inner core (1C) of the Earth are widely assumed to be consistent with those of pure, solid iron in the hexagonal close-packed (ε) phase. This hypothesis is re-examined here using a density model of the Earth's core generated from extrapolated, static high-pressure data for ε iron and pyrite. Densities of constant-composition mixtures at room temperature and core pressures are compared with corresponding Earth-model densities. An effective volume thermal expansivity, α_eff, is calculated for a range of expected core temperatures that brings the room-temperature density into agreement with the Earth-model density at the pressure of the inner-core boundary (ICB). It seems that α_eff would have to be significantly larger than previous estimates of the thermal expansion at core conditions. A qualitatively similar conclusion is obtained if an isotherm reduced from shock-wave data is used for ε iron instead of the static data. We argue that, o several explanations for this difference (errors in Earth-model densities, a high-volume thermal expansivity at megabar pressures, a high-temperature core (>7,000 K), and the presence of a light component), the last alternative is the most probable and that the IC is not, therefore, pure iron.