Two-dimensional (2D) materials have already revolutionized science and have the potential to revolutionize also technology due to their unique properties. 2D semiconducting materials play a particularly important role, as they often combine high carrier mobility with presence of direct band gap and atomic-like thickness.
Properties of 2D materials are affected by perturbations, such as defects and their interaction with other 2D systems via proximity effects in van der Waals (vdW) heterostructures introducing a host of additional complexities, such as formation of Moiré structures, large-scale periodicities, mini-Brillouin zones, variance of spatial profiles and distribution of local environments and properties. Both defects and vdW heterostructures have been studied by computational methods, such as Density Functional Theory (DFT). DFT, while being computationally relatively cheap, may heavily bias both the electronic and structural properties, especially in the vdW hetero layers.
To raise significantly the quality bar, the project team proposes use of the most accurate, albeit also the most computationally demanding method, the ultra-accurate stochastic quantum Monte Carlo (QMC). Due to the length and time-scales involved, cost of brute force application of QMC methods would be prohibitive. They propose a workaround via machine learning tools to study defects in monolayers and twistronics in vdW heterostructures.
Ivan Stich, Institute of Informatics of the Slovak Academy of Sciences, Slovakia