ATOM Lab

Research

Our mission Connecting phenomena across scales is a central challenge in computational materials science. Technologies rely on devices made from many materials working together. We develop and apply electronic structure theory methods to connect their atomic-scale physics with device-scale behavior, helping guide the design of sustainable electronic and photonic technologies, in close collaboration with experimentalists and industry.

Our methods Density-functional theory (DFT), many-body perturbation theory (GW), k·p perturbation theory, and AI/machine learning for discovering equations and effective models. We perform our calculations on the nation’s most powerful supercomputers.

Four hexagonal intensity maps comparing eeh and hhe processes in GaN and AlN.

Theme 01

AI-driven discovery of effective theories for interacting electrons

The properties of materials emerge from how electrons interact with one another and with atomic nuclei. We use AI and machine learning to discover compact equations and effective theories that preserve the essential physics while making these calculations more tractable, allowing us to study more complex systems.

  • Defects for quantum information sciences, electronics, and catalysis.
  • Realistic interfaces for catalysis and electronic devices.

Example publications

Electron and hole distributions in a disordered material, with arrows indicating traps across an 18 nm region.

Theme 02

Designing disorder in materials for next-generation electronics

Disorder changes how electrons move, scatter, and recombine. We develop first-principles methods to predict these effects and understand when disorder limits performance or offers a way to control it.

  • Semiconductor alloys for renewable-energy conversion.
  • Amorphous oxides for next-generation electronics.

Example publications

LED layer structure showing the anode, cathode, p-GaN, AlGaN, quantum well, n-GaN, and GaN buffer.

Theme 03

Connecting quantum chemistry with device physics

Devices contain several materials working together. Understanding them requires connecting the electronic properties of each material to processes across interfaces and throughout the device.

  • Connecting ab initio simulations to TCAD electronic solvers.
  • Designing device heterostructures for LEDs and next-generation AI hardware.

Example publications