Design and Modeling of real-world 2D Materials and Interfaces for AI applications with "Mat3ra-2D"

Computational discovery of 2D materials and interfaces continues to grow in importance due to their applications in microelectronics, energy storage, catalysis, and other critical technology areas [1,2]. We hereby introduce open-source software packages — parts of the the “Mat3ra-2D” stack — architected to work together for end-to-end computational design and modeling of real-world structures. Our approach enables (1) the creation of 2D materials and interfaces from atoms up, including layered materials, van der Waals heterostructures, and interfaces with diverse lattice symmetries and stacking; and (2) inclusion of the effects of strain, surface terminations, defects, disorder and perturbations; (3) the creation and execution of first-principles and AI/ML-driven simulation workflows — including structural relaxations, calculation of properties, such as adhesion energies, electronic band alignment, charge transport, etc. We implement our approach in Python packages — "mat3ra-made", "mat3ra-wode" and others — and include a set of interactive use cases to demonstrate how these tools jointly reproduce and analyze known 2D material systems, providing a data-driven reproducible approach to high-throughput computational design and the corresponding AI/ML applications in materials science.

[1] Shyue Ping Ong, et. al. — Python Materials Genomics (pymatgen) : A Robust, Open-Source Python Library for Materials Analysis. Computational Materials Science, 2013, 68, 314–319.

[2] Ask Hjorth Larsen, et. al.—A Python library for working with atoms J. Phys.: Condens. Matter Vol. 29 273002, 2017