High pressure layered structure of carbon disulfide

TL;DR

This study uses first principles calculations to reveal that high-pressure phases of solid CS₂ are layered structures different from CO₂, with a predicted insulator-metal transition near 50 GPa and implications for superconductivity.

Contribution

It identifies new layered high-pressure structures of CS₂, especially the stable HP1 phase, and predicts their properties, contrasting with previous assumptions based on CO₂.

Findings

HP1 phase is a layered $P2_{1}/c$ structure with edge-sharing tetrahedra.

Predicted Raman spectrum and pair correlation match experiments better than $eta$-cristobalite.

Insulator-metal transition occurs near 50 GPa, but superconductivity origin may differ.

Abstract

Solid CS$_{2}$ is superficially similar to CO$_{2}$, with the same $Cmca$ molecular crystal structure at low pressures, which has suggested similar phases also at high pressures. We carried out an extensive first principles evolutionary search in order to identify the zero temperature lowest enthalpy structures of CS$_{2}$ for increasing pressure up to 200\,GPa. Surprisingly, the molecular $Cmca$ phase does not evolve into $\beta$-cristobalite as in CO$_{2}$, but transforms instead into phases HP2 and HP1, both recently described in high pressure SiS$_{2}$. HP1 in particular, with a wide stability range, is a layered $P2_{1}/c$ structure characterized by pairs of edge-sharing tetrahedra, and theoretically more robust than all other CS$_{2}$ phases discussed so far. Its predicted Raman spectrum and pair correlation function agree with experiment better than those of $\beta$-cristobalite, and further differences are predicted between their respective IR spectra. The band gap of HP1-CS$_{2}$ is calculated to close under pressure yielding an insulator-metal transition near 50 GPa in agreement with experimental observations. However, the metallic density of states remains modest above this pressure, suggesting a different origin for the reported superconductivity.