Combining Artificial Intelligence With Physics Based Simulations in Understanding and Design of Interfaces

Abstract

Interfaces play a critical role in chemical engineering processes, where understanding theirbehavior and properties is essential for process optimization, material design, and technological advancements. However, interfacial problems are inherently complex, characterized by intricate intermolecular interactions and a multitude of influencing factors. To advance our understanding and design of complex interfacial systems, the integration of computational simulations with advanced techniques like machine learning (ML) and data-driven approaches is crucial. This dissertation investigates the challenges and opportunities associated with utilizing simulation and ML techniques to study interfacial phenomena and design chemically patterned surfaces with specific interfacial properties. The research projects presented in this dissertation concentrate on electrode/electrolyte interfaces in lithium-ion batteries, interfacial heat transfer optimization at the Si/Al interface, and the practical application of machine learning potentials (MLPs) for structural relaxation of ternary transition metal dichalcogenides (TMDs). Through these projects, valuable insights are gained into the interplay between surface chemistry and electrolyte dynamics, the optimization of interfacial heat transfer, and the challenges and strategies involved in applying MLPs to custom datasets. This work contributes to the development of an integrated computational workflow, opening avenues for exploring a broader range of interfacial systems, uncovering novel phenomena, and optimizing interfacial properties.