Research

Thermal transport and AI-enabled atomistic simulation

My work connects physics-informed simulation, machine-learning potentials, and open computational tools for understanding heat transport in low-dimensional, interfacial, and complex material systems.

Workflow from DFT-labeled simulation data to MLP-driven molecular dynamics — From DFT-labeled simulation data to deployable machine-learning potentials for large-scale molecular dynamics.

Directions

Research focus

NEP89 chemical space visualization — Machine-learning potential coverage across broad material spaces.

Graphene and hBN thermal transport visualization — Interfacial thermal transport in two-dimensional heterostructures.

Amorphous silica thermal transport visualization — Temperature-dependent transport in amorphous and disordered materials.

Methods

From quantum data to large-scale molecular dynamics

The workflow combines DFT-labeled training data, neuroevolution potentials, and molecular dynamics analysis to bridge accuracy and scale in materials simulation.

Read related publications