Our group focuses on simulating numerically and learning to predict properties of materials in the quantum limit. On one hand, we develop and conduct first principle calculations to study problems of several topics in materials science, such as energy storage and conversion, catalysis, etc. Together with first principle methods, we employ modern technique from data science and machine learning to predict new materials, to search for materials with desired properties, or to solve certain classes of material problems.
On the other hand, we also work on strongly correlated systems, i.e. systems where Coulomb interaction is significant and cannot be neglected and thus valence electrons are correlated and not possible to describe in a picture of a single electron moving in an effective potential. A few examples are Hubbard model, multi-orbital interacting model, transition metal oxides, impurity problems. Non-equilibrium correlated systems are also under investigation, e.g. studying the time evolution of Kondo resonance in response to a sudden quench and the possibility to observe such time-evolution in experiments.