The iron-sulfur system is important for planetary interiors and is intensely studied, particularly for better understanding of the cores of the Earth and of terrestrial planets. Yet, there is a paradox about high-pressure stability of FeS: ab initio global optimization (at DFT level) predicts a $Pmmn$ phase (with a distorted rocksalt structure) to be stable at pressures above $\sim 120$ GPa, which has not yet been observed in the experiments, which instead revealed a CsCl-type phase which, according to density functional calculations, should not be stable. Using quasiharmonic free energy calculations and dynamical mean-field theory, we show that this apparent discrepancy is removed by proper account of electron correlations and entropic effects.

• Received 2 November 2017
• Revised 17 August 2018

DOI:https://doi.org/10.1103/PhysRevB.98.094112