Abstract
Graphene and 2D materials, despite being relatively fresh materials, have already taken a firm place in research, development and applications. A number of exciting phenomena have been discovered in these crystals and they continue bringing exciting results on a regular basis. However, probably the most important characteristic about 2D materials is that they offer a possibility to form on-demand van der Waals heterostructures, where individual 2D crystals are stacked together, forming a novel, 3D structure, which composition (and thus, their properties) can be controlled with atomic precision. This has opened a new direction of research: materials on demand. The properties of the resulting heterostructure can be designed with very high precision. The space of parameters is so large that the use of machine learning methods becomes essential. Furthermore, since individual components in such heterostructures interact through a number of channels (elastic, van der Waals, electronic, etc.) – a degenerate energy landscape is formed, leading to a number of competing phases, which opens a way to engineer particular phase transitions between different states and, thus, study also the out-of-equilibrium phenomena in such structures.
