Fig. 5: Generalization ability of DeepH, from flat sheets to curved nanotubes. | Nature Computational Science

Fig. 5: Generalization ability of DeepH, from flat sheets to curved nanotubes.

From: Deep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculation

Fig. 5: Generalization ability of DeepH, from flat sheets to curved nanotubes.

a, The use of DeepH, trained by DFT results for flat sheets, to study curved nanotubes: graphene (left) and MoS2 (right). b,c, Band structures for a zigzag (25, 0) carbon nanotube (b) and a zigzag (50, 0) MoS2 nanotube (c), computed by DFT and DeepH. The Fermi level is aligned at the middle of the bandgap.

Source data

Back to article page