Executive Summary : | The structure and morphology, determined by self-assembly, significantly impact the material's properties. Therefore, our ability to control the properties of a material very much depends on our ability to control structure at desired length scales. Furthermore, recent experiments suggest that the self-assembly processes in the presence of an interface or confinement are strikingly distinct from those observed in bulk systems. The ability to control the self-assembly process in the presence of an interface (solid-fluid) and confinement has opened up new possibilities for the design and synthesis of functional materials with exotic properties. However, despite much current interest, there is a significant gap in establishing a direct relationship between the chemistry of the surface (or, the confining wall) and pathways and kinetics of self-assembly of the nanoscale particles. One major challenge lies in unraveling the non-trivial connection between the interface-induced changes in the solvent and solvent-mediated effective interaction between the particles. The interface-induced spatio-temporal heterogeneities in the solvent can alter the solvent-mediated interactions between the particles, and in turn, the self-assembly pathways. In recent years, advancements in simulation techniques and resources have made it more practical to use computational modeling to design and study functional materials using bottom-up approaches. In this project, we propose to address four main aspects: (i) characterization of surface-induced solvent's (short and medium-long-range) structural heterogeneities using persistence homology (a topological data analysis method) and machine learning, (ii) the interplay between the surface-induced structural heterogeneities in the solvent and solvent-mediated surface-particle and particle-particle \textit{effective} interactions, (iii) the dependence of the self-assembly pathways and kinetics on the surface chemistry, and (iv) computational and machine learning investigation of the sensitivity of nucleation propensity of a surface on its (surface) chemistry for soft colloids. Applying machine learning techniques enables a physically motivated input to the model that can be computed or experimentally measured. Thus, this project aims to provide a general framework to predictably control the self-assembly pathways of nanoscale building blocks through interface-driven nucleation to design the desired structure. The predictive control of the self-assembly pathways has tremendous practical relevance in fields of science and technology as diverse as materials and biological sciences. |