Advanced Material Design: Multiscale Modeling and Control of Nanostructure-Property Relationships in Self-Assembling Amphiphiles

This study addresses the development and application of reversible complex fluids through a detailed examination of the self-assembled nanostructures that dictate their properties. Reversible complex fluids are pivotal in various industries, including pharmaceuticals, oil and gas, and specialty chemicals, where the reversibility of these fluids is chiefly attributed to dynamic molecular interactions within self-assembled nanostructures composed of amphiphilic molecules. These molecules, which feature distinct polar heads and non-polar tails, spontaneously form various structural geometries such as bilayers, spherical micelles, and wormlike micelles (WLMs). Such structures are essential, as they significantly influence the fluids’ viscoelastic properties, with WLMs providing enhanced characteristics due to their entanglement capabilities compared to the simpler spherical micelles. Our research synthesizes numerous theoretical insights to link the molecular structure of amphiphiles directly to the resulting nanostructures, and subsequently to the macroscopic properties of the fluids. By leveraging advanced simulation techniques such as molecular dynamics and density functional theory, coupled with empirical data, we establish a robust framework that not only describes nanostructure formation but also facilitates the manipulation of these structures through controlled adjustments in amphiphile chemistry and reaction conditions. The resulting framework offers a systematic approach to optimize the design and application of complex fluids, incorporating multiscale modeling to bridge molecular interactions with large-scale fluid behavior. This integration proves crucial for practical applications such as hydraulic fracturing, where the specific properties of the fluid significantly impact performance. The study ultimately provides a comprehensive strategy for predicting and enhancing the on-field effectiveness of complex fluids, informed by a deep understanding of their molecular foundations and controlled through advanced modeling and empirical practices. The development of physics-based models for material design often encounters challenges due to the inherent uncertainties of first-principles models, which affect their utility and accuracy, especially in on-field performance analysis and decision-making. To mitigate these issues, we introduce a hybrid modeling framework that combines the interpretability of first-principles models with the adaptability of advanced machine learning techniques, including neural networks. This framework focuses on reducing uncertainties in parameters that vary over time or space, thereby enhancing model predictiveness. It features a novel branched architecture of convolutional and deep neural networks to manage spatiotemporal uncertainties effectively. A pivotal case study shows the framework’s ability to accurately estimate dynamic parameters in reaction-diffusion systems, improving prediction accuracy and uncovering hidden chemical dynamics. Additionally, we implement a physics-informed regularization scheme to address solution multiplicity challenges, increasing the robustness of our approach. With full validation, this model will be applied to predict key properties such as viscoelasticity and critical micelle concentration, significantly advancing the analysis and management of complex fluids in sectors like hydraulic fracturing.

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Chemical Engineering

Grantor

Texas A&M University

Year dc:date.issued

2024

Author dc:creator

• Pahari, Silabrata

Advisor dc:contributor.advisor

• Kwon, Joseph Sang-Il

Committee members dc:contributor.committeemember

• Kravaris, Costas
• Gildin, Eduardo
• Akbulut, Mustafa

• Engineering, Chemical
• Engineering, Materials Science
• Energy

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/1969.1/1591514

OAI record oai:identifier

https://oaktrust.library.tamu.edu/server/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Aoaktrust.library.tamu.edu%3A1969.1%2F1591514