Search

Crystal structure design is important for the discovery of new highly functional materials because crystal structure strongly influences material properties. Crystal structures are composed of space-filling polyhedra, which affect material properties such as ionic conductivity and dielectric constant. However, most conventional methods of crystal structure prediction use random structure generation methods that do not take space-filling polyhedra into account, contributing to the inefficiency of materials development. In this work, we propose a crystal structure generation method based on discrete geometric analysis of polyhedra information. In our method, the shape and connectivity of a space-filling polyhedron are represented as a dual periodic graph, and the crystal structure is generated by the standard realization of this graph. We demonstrate that this method can correctly generate face-centered cubic, hexagonal close-packed, and body-centered cubic structures from dual periodic graphs. This work is a first step toward generating undiscovered crystal structures based on the target polyhedra, leading to major advances in materials design in areas including electronics and energy storage.

Entangled objects such as entangled graphs and weaves are often seen in nature. In the present article, two identical graphs entangled, and weaves with two different color threads are studied. A method to identify stable configurations in the three dimensional space of a given topological entangled structure, a planar graph with crossing information, is proposed by analyzing the steepest descent flow of the energy functional with repulsive interactions. The existence and uniqueness of a solution in the entangled case. In the untangled case, a weave has a unique tangle decomposition with height order, whose components are moving away each other in the order $t^{1/3}$ as time $t$ goes to the infinity., Comment: 37 pages, 4 figures

A discrete harmonic surface is a trivalent graph which satisfies the balancing condition in the 3-dimensional Euclidean space and achieves energy minimizing under local deformations. Given a topological trivalent graph, a holomorphic function, and an associated discrete holomorphic quadratic form, a version of the Weierstrass representation formula for discrete harmonic surfaces in the 3-dimensional Euclidean space is proposed. By using the formula, a smooth converging sequence of discrete harmonic surfaces is constructed, and its limit is a classical minimal surface defined with the same holomorphic data. As an application, we have a discrete approximation of the Enneper surface.

We find that simple neural networks with ReLU activation generate polytopes as an approximation of a unit sphere in various dimensions. The species of polytopes are regulated by the network architecture, such as the number of units and layers. For a variety of activation functions, generalization of polytopes is obtained, which we call neural polytopes. They are a smooth analogue of polytopes, exhibiting geometric duality. This finding initiates research of generative discrete geometry to approximate surfaces by machine learning., Comment: 5 pages, 9 figures. v2: References added. Accepted at the 1st Workshop on the Synergy of Scientific and Machine Learning Modeling at International Conference on Machine Learning (ICML), Honolulu, Hawaii, USA. 2023

We propose a method to calculate wave functions and energies not only of the ground state but also of low-lying excited states using a deep neural network and the unsupervised machine learning technique. For systems composed of identical particles, a simple method to perform symmetrization for bosonic systems and antisymmetrization for fermionic systems is also proposed., Comment: 23 pages, 18 figures, 8 tables; Typo of Eq. (25) corrected