This article introduces DeepH-hybrid, a neural network approach designed to enhance the efficiency of hybrid density functional theory (DFT) calculations, which are critical for accurate material predictions. This method overcomes the computational challenges of hybrid functionals by using deep equivariant neural networks that learn the relationship between material structures and their electronic properties. DeepH-hybrid offers hybrid-functional accuracy while significantly reducing computation time, allowing for large-scale material simulations, such as studying complex systems like twisted bilayer graphene and transition metal dichalcogenides.
Please continue reading the full article under the following link:
https://www.nature.com/articles/s41467-024-53028-4
#materials #materialsscience #materialsengineering #computationalchemistry #modelling #chemistry #researchanddevelopment #research #MaterialsSquare #ComputationalChemistry #Tutorial #DFT #simulationsoftware #simulation