*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
THE SCHOOL OF MATERIALS SCIENCE AND ENGINEERING
GEORGIA INSTITUTE OF TECHNOLOGY
Under the provisions of the regulations for the degree
DOCTOR OF PHILOSOPHY
on Thursday, April 20, 2022
10:30 AM
via
BlueJeans Video Conferencing
https://bluejeans.com/503834048/2526
will be held the
DISSERTATION PROPOSAL DEFENSE
for
Yifan Liu
"Design of Organic-Inorganic Hybrid Membranes Using Density Functional Theory and Machine Learning"
Committee Members:
Prof. Rampi Ramprasad, Advisor, MSE
Prof. Mark D. Losego, MSE
Prof. Roshan V. Joseph, ISYE
Prof. Ryan P. Lively, ChBE
Prof. Zhiqun Lin, MSE
Abstract:
Vapor phase infiltration (VPI) has the ability to produce organic-inorganic hybrid membranes with enhanced stability in organic solvents, while retaining high permeance and selectivity. Due to the vastness of the design space for such membranes, which includes polymer chemistry, inorganic chemistry, and hybrid microstructures, traditional trial and error materials development methods are inefficient. To address this challenge, this work outlines three objectives to develop the knowledge and tools to predict and explore novel VPI organic-inorganic membranes using density functional theory (DFT) and machine learning (ML). In Objective 1, the interactions between three representative metal precursors and a prototype polymer, polymer of intrinsic microporosity 1 (PIM-1), during the VPI process have been studied by DFT. In Objective 2, an ML model was developed to predict the sublimation enthalpy of metal precursors and assist the chemistry selection and experimental process design. Due to a paucity of literature-reported values in metal precursor parameters, a DFT dataset of organic molecules was first generated to train the ML model, and an active learning algorithm was subsequently developed to include metal precursors in the model. To extend the studies in Objective 1, an ML model with a newly generated DFT dataset incorporating multi-task learning will be developed as part of Objective 3, to predict binding energy between metal precursors and polymers. The above predictive models, along with the chemical guidelines obtained from feature analysis, will aid the selection of potential candidates of polymers and metal precursors in designing new organic-inorganic hybrid membranes for energy-efficient chemical separation.
