Laboratory of Digital Material Science

Various first-order phase transitions in their development pass through the same stages, the first of which is the nucleation stage, which is of the greatest interest and greatest difficulty in studying. In the theory of this stage, questions of the thermodynamics of small systems and the description of the process of overcoming the energy barrier by nascent particles are closely intertwined. To achieve a detailed understanding of nucleation, it is necessary to use computer modeling methods. The small size of the nascent core of the new phase requires taking into account the contributions of the interface, surface energy, relaxation of mechanical stresses into the curvature energy and other features of low-dimensional materials. This requires high-precision modeling that takes into account all these parameters, which, however, is an extremely difficult task for the current tools of computational materials science. Indeed, traditional methods of electron density functional theory, although they make it possible to quite accurately calculate the properties of atomic systems from first principles, are nevertheless limited by the available computing power. This limits their applicability to periodic structures consisting of hundreds of atoms. At the same time, the task of describing the nucleation of new phases requires the description of systems with the number of atoms up to 10^4-10^6 atoms. On the other hand, empirical potentials that are not demanding on computational resources make it possible to describe large systems containing millions of atoms. But until recently, the parameterization of these potentials was limited to their own (often rather narrow) model systems, which were not intended for modeling transition states and new phases, which is a necessary condition for the study of phase transformations. However, the situation has changed dramatically recently with the advent of empirical machine learning capabilities that can be trained on large data sets obtained through first-principles calculations. Thus, one of the objectives of the project is to develop such potentials that describe the interaction with precision methods from first principles, allowing one to model the required number of atoms in structures. Parameterized potentials will be used for the scientific study of phase transformation in carbon systems, the graphite-diamond transition and multilayer graphene-ultrathin diamond film (diamane), and in the future can be used to describe phase transitions in other systems and nanomaterials.