Fig. 2
From: Predicting superhard materials via a machine learning informed evolutionary structure search
The hardness vs. energy map of all carbon topologies predicted by the XtalOpt searches. The horizontal line corresponds to a \(H_{{\mathrm{v,Teter}}}^{{\mathrm{ML}}} = 40\) GPa, and the vertical line to an energy of −8.80 eV/atom. The structures in the yellow quadrant are both stable and superhard
