A previously unknown type of elemental structure has been discovered recently in barium, phase IV. The structure is composed of a “host” and two or more “guest” components made up of chains that lie in channels in the host, and form structures that are incommensurate with the host. This implies a segregation into different electronic species. The same structure has now been found in phase V of Sr above 45 GPa.

An SC16 phase of GaAs has been observed at high pressure. The phase is obtained by heating the high-pressure $\mathrm{Cmcm}$ phase to above $\sim 400\mathrm{K}$ at $\sim 14\mathrm{GPa}$ and is found to be stable on cooling back to room temperature. The detailed structure has been refined. On pressure decrease, the SC16 phase is found to be stable down to atmospheric pressure. From high-temperature studies, the equilibrium range for the SC16 structure is 13 to 14.5 GPa.

The crystal diffusion variational autoencoder (CDVAE) is a machine learning model that leverages score matching to generate realistic crystal structures that preserve crystal symmetry. In this study, we leverage novel diffusion probabilistic (DP) models to denoise atomic coordinates rather than adopting the standard score matching approach in CDVAE. Our proposed DP-CDVAE model can reconstruct and generate crystal structures whose qualities are statistically comparable to those of the original CDVAE. Furthermore, notably, when comparing the carbon structures generated by the DP-CDVAE model with relaxed structures obtained from density functional theory calculations, we find that the DP-CDVAE generated structures are remarkably closer to their respective ground states. The energy differences between these structures and the true ground states are, on average, 68.1 meV/atom lower than those generated by the original CDVAE. This significant improvement in the energy accuracy highlights the effectiveness of the DP-CDVAE model in generating crystal structures that better represent their ground-state configurations.