3700 O'Hara Street, Pittsburgh, PA 15261

 

 PhD Dissertation Final Presentation 

PhD Computational Modeling and Simulation Engineering Major,

PhD Chemical Engineering Concentration

 

Molecular Modeling with Atomistic Machine Learning Methods

 

Abstract:

Simulations using quantum-mechanics (QM) methods offer insights into ion and atom diffusion, transition states of reactions, and molecular electronic structures, with applications in engineering porous materials, fuel cell membranes, and molecular characterization. However, their scope is limited by computational constraints, restricting studies to smaller systems and shorter time scales. To address this, atomic configurations can be mapped to QM data using efficient machine learning (ML) models.

 We focus on developing ML forcefields for various applications. Firstly, we trained highly accurate deep learning potentials (DPs) for graphanol (hydroxy functionalized graphane). Our simulations demonstrated that graphanol conducts protons efficiently without hydration. Our investigations into proton diffusion and barriers, along with temperature fluctuations, revealed insights for designing improved proton exchange membranes. Additionally, we employed accurate DPs for modeling diffusion in metal-organic frameworks (MOFs) like UiO-66, and interface diffusion in chalcogenide alloys and electrodes for non-volatile memory cells, using the moment tensor approach (MTP) for construction. Training these forcefields typically relies on molecular dynamics (MD)-based active learning, which is inefficient for accurately predicting chemical reactions. To overcome this, we developed a reactive active learning approach that automates reaction generation and employs transition-state finding techniques. This active learning scheme resulted in accurate prediction of reaction barriers with fewer configurations compared to traditional MD-based active learning.

Committee: 

Dr. J. Karl Johnson (Chair, Department of Chemical and Petroleum Engineering)

Dr. John Keith (Department of Chemical and Petroleum Engineering)

Dr. Leonardo Bernasconi (Center for Research Computing & Department of Chemistry)

Dr. Geoffrey Hutchison (Department of Chemistry)

Dr. John Kitchin (Department of Chemical Engineering, Carnegie Mellon University

Dr. Derek Stewart (Western Digital)

 

Event Details

Please let us know if you require an accommodation in order to participate in this event. Accommodations may include live captioning, ASL interpreters, and/or captioned media and accessible documents from recorded events. At least 5 days in advance is recommended.


Link: https://pitt.zoom.us/my/sidachar 

Meeting ID: 450 354 3543

Passcode: w1DME8

University of Pittsburgh Powered by the Localist Community Event Platform © All rights reserved