Observed Metallization of Hydrogen Interpreted as a Band Structure Effect

A recent experimental study of the metallization of hydrogen tracked the direct band gap and vibron frequency via infrared measurements up to ~425 GPa [P. Loubeyre et al., Nature 577, 631 (2020)]. Above this pressure, the direct gap has a discontinuous drop to below the minimum experimentally accessible energy (~0.1 eV). The authors suggested that this observation is caused by a structural phase transition between the C2/c-24 molecular phase to another molecular phase such as Cmca-12. Here, through ab initio calculations of pressure dependent vibron frequency and direct band gap, we find that the experimental data is consistent with the C2/c-24 phase up to 425 GPa, and suggest that this consistency extends beyond that pressure. Specifically, we find that qualitative changes in the band structure of the C2/c-24 phase lead to a discontinuous drop of the direct band gap, which can explain the observed drop without a structural transition. This alternative scenario naturally explains the absence of hysteresis in the measurements.