High-pressure behavior of dense hydrogen up to 3.5 TPa from density functional theory calculations

TL;DR

This study uses density functional theory to analyze the structural behavior and equation of state of dense hydrogen up to 3.5 TPa, revealing new phases, transition barriers, and limitations of low-pressure models.

Contribution

It provides a comprehensive DFT-based analysis of high-pressure hydrogen phases, correcting previous misconceptions and exploring transition pathways and energy barriers.

Findings

Vinet EOS overestimates compressibility beyond 500 GPa

Metastable multi-atomic phases appear above 2.8 TPa

Transition barriers can raise dissociation pressures by about 200 GPa

Abstract

Structural behavior and equation of state of atomic and molecular crystal phases of dense hydrogen at pressures up to 3.5 TPa are systematically investigated with density functional theory. The results indicate that the Vinet EOS model that fitted to low-pressure experimental data overestimates the compressibility of dense hydrogen drastically when beyond 500 GPa. Metastable multi-atomic molecular phases with weak covalent bonds are observed. When compressed beyond about 2.8 TPa, these exotic low-coordinated phases become competitive with the groundstate and other high-symmetry atomic phases. Using nudged elastic band method, the transition path and the associated energy barrier between these high-pressure phases are evaluated. In particular for the case of dissociation of diatomic molecular phase into the atomic metallic Cs-IV phase, the existent barrier might raise the transition pressure about 200 GPa at low temperatures. Plenty of flat and broad basins on the energy surface of dense hydrogen have been discovered, which should take a major responsibility for the highly anharmonic zero point vibrations of the lattice, as well as the quantum structure fluctuations in some extreme cases. At zero pressure, our analysis demonstrates that all of these atomic phases of dense hydrogen known so far are unquenchable. NOTE: In the previous version of this paper (1010.3392v1) we made a mistake when evaluating the enthalpy of Cs-IV phase, which misled us to a conclusion that taking the multi-atomic molecular phases as the ground-state. After corrected this error, however, the atomic phase of Cs-IV becomes the static structure with the lowest enthalpy. Current version not only includes a substantial improvement of the previous one, but also contains many NEW interesting topics that were not touched before.