Molecular Dynamics for Semiconductor Phonon and Thermal Dynamics

Fri, 29 May 2026 00:00:00 +0000

Project Mission

This project develops molecular dynamics methods for understanding semiconductor materials at finite temperature. We focus on how atomic motion, lattice vibrations, and phonon damping influence thermal transport, optoelectronic stability, and device-relevant material behavior.

The goal is to build Hamiltonian- and DFT-informed simulation workflows that connect atomistic trajectories with semiconductor performance, especially for perovskites and optoelectronic materials.

Scientific Motivation

Semiconductor behavior is not determined only by static band structures. Real materials operate under temperature, strain, illumination, interfaces, defects, and structural disorder. These conditions create dynamic atomic environments where phonon scattering, lattice anharmonicity, carrier relaxation, defect migration, and interfacial heat transfer can strongly affect semiconductor device performance.

Molecular dynamics provides a direct way to study these time-dependent processes. Instead of only predicting isolated material properties, this project investigates how semiconductor structures evolve, dissipate energy, and respond to thermal perturbations over time.

Research Approach

Hamiltonian-Informed Molecular Dynamics

Ab Initio Molecular Dynamics: Using DFT-based forces to simulate atomic motion without relying only on empirical force fields.
Hamiltonian Modeling: Constructing electronic-structure-aware representations that connect atomic configurations with energy landscapes, forces, and dynamical responses.
Born-Oppenheimer Dynamics: Simulating nuclear trajectories on electronic ground-state potential energy surfaces for finite-temperature materials analysis.
Car-Parrinello Dynamics: Exploring coupled electron-ion propagation for efficient first-principles molecular dynamics.
Trajectory-Based Electronic Analysis: Evaluating how band edges, local structure, bonding environments, and defect states fluctuate along atomic trajectories.

Phonon Dynamics and Damping

Phonon Lifetime Analysis: Estimating how long vibrational modes persist before scattering or dissipating energy.
Phonon Damping in Semiconductors: Studying damped lattice vibrations caused by anharmonicity, defects, disorder, and electron-phonon interactions.
Thermal Conductivity Modeling: Connecting phonon transport and heat dissipation to semiconductor device reliability.
Anharmonic Lattice Dynamics: Modeling non-harmonic atomic motion that becomes important at finite temperature and under structural instability.
Hot-Carrier and Energy Relaxation: Investigating how lattice vibrations mediate energy loss in photovoltaic materials.

Semiconductor and Perovskite Materials

Halide Perovskites: Studying soft lattice motion, ion migration, octahedral tilting, phonon broadening, and thermal instability.
Optoelectronic Semiconductors: Modeling materials used in solar cells, LEDs, photodetectors, and light-emitting devices.
Defect Dynamics: Simulating how vacancies, interstitials, grain boundaries, and local disorder evolve under thermal conditions.
Interface Stability: Investigating atomic-scale behavior at semiconductor interfaces, heterostructures, and contact layers.
Strain and Thermal Response: Evaluating how mechanical deformation changes phonon transport and lattice stability.

AI-Accelerated Molecular Dynamics

Machine Learning Interatomic Potentials: Training neural potentials from DFT data to extend molecular dynamics to larger systems and longer timescales.
Active Learning for Force Fields: Selecting informative atomic configurations for new DFT calculations during simulation.
Uncertainty-Aware Trajectories: Detecting when a learned potential leaves its reliable domain during semiconductor dynamics.
Graph-Based Atomic Representations: Encoding local atomic neighborhoods for transferable force and energy prediction.
Multiscale Simulation: Bridging first-principles accuracy with device-relevant length and time scales.

System Design

The proposed workflow begins with DFT and ab initio molecular dynamics calculations for representative semiconductor structures. These simulations generate atomic trajectories, forces, stresses, local electronic features, and finite-temperature structural information.

The next stage trains machine learning interatomic potentials and Hamiltonian-aware models from first-principles data. These models enable longer molecular dynamics simulations for phonon transport, damping behavior, defect motion, and interface evolution.

The analysis layer extracts physical quantities such as phonon spectra, mode lifetimes, thermal conductivity, energy dissipation pathways, defect migration barriers, and structural stability indicators. This creates a simulation pipeline that connects atomistic dynamics with semiconductor material design.

Current Implementation

At the current stage, this project focuses on building molecular dynamics workflows for semiconductor lattice dynamics and thermal transport. The initial implementation uses DFT calculations, ab initio molecular dynamics, classical molecular dynamics, and phonon analysis tools to study finite-temperature behavior in semiconductor systems.

The near-term focus is on perovskite and optoelectronic materials, where soft lattice motion, phonon damping, defect dynamics, and electron-phonon interactions are critical to material stability and performance.

Future Research Directions

Future work will extend this project toward Hamiltonian-informed and AI-accelerated molecular dynamics. One direction is to develop learned DFT Hamiltonian models that can approximate electronic-structure information along molecular dynamics trajectories more efficiently than repeated self-consistent DFT calculations.

Another direction is to combine molecular dynamics with nonadiabatic and electron-phonon coupling models to study carrier relaxation, hot-carrier cooling, and phonon-mediated recombination in semiconductors.

The long-term objective is to build a semiconductor dynamics platform that can predict how materials behave under realistic operating conditions, including temperature gradients, illumination, defects, interfaces, and mechanical strain. This platform will support the design of more stable perovskites, efficient optoelectronic materials, and thermally robust semiconductor devices.

Phonon Damping | Research Lab