Low Dimensional Materials
We are interested in low dimensional materials, including their electronic
properties, magnectic properties and superconductivity. These materials have been found to have applications such as the next generation of nanoscale semiconductor devices,
solar cells and chemical catalyst. Recently, the study of graphene-like materials such
as silicene, germanene, borophene and monolayer boron-carbon have aroused great interest
among researchers in the field of condensed matter physics.Besides, combined with statistical physics method, we have the ability to study electronic properties of nanostructures, the metastable state and dynamics of growth processes.
Phase Transitions of Materials
Phase diagrams are the foundation in performing basic materials research and can serve as a road map for materials design and process optimization. Calculation of phase diagrams is of great interest since it helps understanding the mechanics of phase stability and phase transition. The difficulty in solve the phase diagram numerically is to find a proper model that describes the system clearly and precisely. Now the well-developed phase diagrams calculations methods are cluster variation method (CVM), CALPHAD, and Monte Carlo simulations. As the development of computational technique and the requirement of materials design, the calculation of alloy phase diagram has expanded from binary, ternary alloy to high entropy alloys (HEAs).
Density functional theory(DFT) is a mature technique for calculating the
structure and behavior of solids, which enables the development of extensive databases that cover the calculated properties of known and hypothetical systems. The data from experiments and High-throughput calculations constitute mature materials' databases. Combing adequate data and the machine learning algorithm, we could obtain a powerful tool for screening and material design. The choice of different machine learning model and the descriptors of materials are key factors in the process of predicting material properties. We construt different models to predict the properties of inorganic materials. The models are trained by data from those databases. By the well-trained models, we can find new materials with good performance, which could be validated by the first-principles calculations.