Revised shock-wave equations of state for high-pressure phases of rocks and minerals

Abstract

Shock‐wave data for the high‐pressure phases of a number of rocks and minerals have been reanalyzed using a revised seismic equation of state to constrain the zero‐pressure properties of the high‐pressure phases. The anomalously low values of dK/dP resulting from a previous analysis are thereby removed. The inferred zero‐pressure densities of the high‐pressure phases are reduced by an average of 4%, and the values of the zero‐pressure seismic parameter Φ_0 are reduced by up to 30%. bringing them into approximate agreement with the hypothesis of the molar additivity of Φ. For most of the materials considered, the derived pressure trajectories of density versus the seismic parameter Φ are consistent wth shock‐wave data on such materials as MgO, Al_2O_3, and SiO_2 (stishovite) where no zero‐pressure assumptions are required. Iron‐rich compounds may require further revision. It is demonstrated that in poorly constrained cases the Birch‐Murnaghan equation can produce a singularity in dK/dP at high pressure. Possible crystal structures of the high‐pressure phases are considered using the revised zero‐pressure densities. It seems likely that olivines with less than about 10‐mole % FeO content can transform to a phase significantly denser than the isochemical mixture of oxides, in contrast to olivines with higher iron content. The possibility that electron spin transitions occur in iron‐rich compounds is considered, but no strong evidence has been obtained. The derived zero‐pressure densities of the high‐pressure phases are usually not of sufficient accuracy to distinguish between all alternative structures, but in some cases an alternative structure to that previously chosen is preferred here.