Overhauser [Phys. Rev. B 35, 411 (1987)] has recently proposed that LiBe${\mathrm{H}}_3$ is a possible high-$T_c$ superconductor. A crucial question is whether LiBe${\mathrm{H}}_3$ is metallic. To address this question, we have employed the pseudopotential total-energy method to investigate the properties of LiBe${\mathrm{H}}_3$. Two cubic perovskite structures are considered, one with Be at the cube center and Li at the cube corner and the other with their positions reversed. The former is a semiconductor with a calculated band gap of about 0.8 eV while the latter is a metal. The semiconductor structure is found to be more stable by 0.77 eV/atom. The equilibrium lattice constants of the two structures are almost the same, ∼3.18 Å, yielding a density which is about half that of the structure proposed by Overhauser. Possible methods to achieve a metallic phase are explored. In particular, our results indicate that at high pressure ($P\simeq 0.7$ Mbar) the semiconducting structure metallizes and remains more stable compared to the other structure.

• Received 5 April 1988

DOI:https://doi.org/10.1103/PhysRevB.38.3576