William Schertzer

Advisor: Prof. Rampi Ramprasad

will propose a doctoral thesis entitled,

AI-Guided Investigation of Polymers for The Design of Robust Anion Exchange Membrane Fuel Cells

On 

Thursday, Nov. 20, 2025

10 am - 12 pm 

In 

MRDC Room 3515

or 

virtually via Teams:

Committee:

Prof. Rampi Ramprasad- School of Materials Science and Engineering (advisor)

Prof. Ryan P. Lively- School of Chemical and Biomolecular Engineering

Prof. Chao Zhang- School of Computational Science and Engineering

Prof. Scott Danielsen- School of Materials Science and Engineering

Prof. Guoxiang (Emma) Hu- School of Materials Science and Engineering

Abstract: As the global demand for sustainable energy continues to rise, polymer-based anion exchange membranes (AEMs) have emerged as a promising platform for next-generation fuel cells that operate under alkaline conditions. However, the development of high-performance and durable AEMs is hindered by the vast design space of possible chemistries, the trade-offs among key transport and mechanical properties, and the scarcity of high-quality, structured experimental data. This thesis aims to accelerate the discovery, understanding, and lifetime prediction of AEM materials through a data-driven framework that integrates machine learning, physics-based modeling, and automated knowledge extraction from the scientific literature. The first part of this work establishes a computational pipeline for novel AEM copolymer design, where predictive models trained on curated literature data identify fluorine-free candidates with optimal combinations of hydroxide conductivity, water uptake, and swelling ratio. The second part introduces a physics-enforced neural network (PENN) that learns universal degradation behavior across diverse AEM chemistries and operating conditions, enabling the forecasting of long-term conductivity decay (up to 10,000 h) from minimal early-time data. The final part of the thesis leverages optical character recognition, computer vision, large language models, and heuristics to automate the extraction of complex, context-rich data from figures, schematics, tables, and text within AEM literature. Together, these efforts will create a closed-loop platform for polymer discovery and degradation modeling, transforming how experimental knowledge is captured and applied to accelerate the design of sustainable, high-performance materials for clean energy technologies.