Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Chemical and Biomolecular Engineering

Committee Chair/Advisor

Sapna Sarupria

Committee Member

Eric Davis

Committee Member

Rachel Getman

Committee Member

David Ladner


Developing fouling-resistant membranes is essential for providing affordable access to clean water for the growing global population. Membrane technology offers an efficient and environment-friendly approach to producing large quantities of clean water through seawater desalination. Additionally, extracting valuable resources like lithium from brine using ion-selective membranes enhances the economic viability of desalination plants. However, challenges such as declining membrane performance due to fouling and limited ion selectivity hinder the broader affordability of membrane-based desalination and resource recovery. We use molecular dynamics (MD) simulations to assess the fouling mechanism and selectivity of polyamide membranes.

One of the key challenges in studying polyamide membranes using molecular simulations is the lack of an accurate all-atom structure. MD simulations rely on precise all-atom structures and parameterization for reliable estimates of material properties. To address this, we develop a generalized software, PolymerizeIt!, which facilitates the generation of all-atom structures of cross-linked polymers. This software allows users to create the polyamide structure and demonstrates its extensibility on two other cross-linked polymers. PolymerizeIt! enables users to write customized polymerization protocols using intuitive Python scripts, enhancing the flexibility of integrating new features into various polymerization protocols.

Understanding the influence of membrane chemistry on interfacial hydration and fouling propensity at the nanoscale is pivotal for making informed design decisions regarding fouling-resistant membranes and reducing reliance on heuristic approaches. We investigate the thermodynamic origins of the fouling phenomenon of polyamide reverse osmosis membranes, which are widely used in commercial seawater desalination. Specifically, the interfacial water structure and foulant-membrane interactions are analyzed to understand the interfacial hydration and fouling mechanism. Several small molecules with varying polarities are examined as model foulants. The overall surface is hydrophilic, although local hydrophobic patches and some hydrophobic signatures in interfacial water are also observed. Fouling on polyamide membranes is driven more by hydrophilic than hydrophobic interactions. Drawing on findings from studies on model Self-Assembled Monolayer (SAM) surfaces, we evaluate potential improvements in fouling resistance by implementing a pattern of alternating polar and non-polar moieties.

While the studies on interfacial water and interactions between membranes and small solutes provide useful insights into the fouling mechanism, confirming these trends by studying larger foulants representing realistic biomolecules and organic molecules is crucial. However, enhanced sampling techniques are required to study protein fouling within tractable simulation timeframes. Enhanced sampling techniques depend on selecting an appropriate collective variable (CV), which is not intuitive except for very simple systems. We utilize a data-driven method called MESA to study protein adsorption on polyamide, conducting a proof-of-concept study of the adsorption of the trp-cage mini protein by comparing the free energy surface obtained from MESA to that from umbrella sampling along known CVs.

For enhancing ion-ion selectivity, an improved understanding of the combined effects of nanoporous structure and chemistry of membrane on ion and water transport through the membranes is essential. However, ion transport through the membranes is a rare event. The enhanced sampling techniques utilized to study ion transport and selectivity mechanisms are expensive and impede rapid exploration of design space with membrane structure and chemistry. We employ a high-throughput method that utilizes an electric field gradient to drive ions through the membrane. We obtain measurable ion transport through polyamide nanofiltration membranes within tractable simulation time in unbiased MD trajectories. We provide a proof-of-concept of the method by validating known selectivity trends between monovalent and divalent ions.

Overall, this dissertation focuses on improving the understanding of the fouling mechanism and selectivity of polyamide membranes using molecular dynamics simulations and developing the tools required to conduct these investigations.

Available for download on Saturday, May 31, 2025