-
Welcome I-PhD project student Yuvraj Singh! (December 2019)
-
Welcome I-PhD project student Yuvraj Singh! (December 2019)
Phase Transition in Equilibrium and Driven Out-of-Equilibrium Systems
The formation of many of the complex hierarchical structures in nature involves either complex/rugged free energy landscape (e.g., zeolites, proteins, etc.) or driven out-of-equilibrium conditions (e.g., intra-cellular self-organization of cytoskeletons, birds flocking, etc.). These complex self-organized structures play key functional roles in many of the physical, chemical and biological processes in nature. Understanding the principles governing the self-assembly and self-organization in these complex (equilibrium and driven out-of-equilibrium) systems remains one of the central challenges of statistical mechanics. Using computational modeling and theoretical tools rooted in statistical mechanics we are interested in gaining fundamental understanding of self-assembly processes involving complex free energy landscape at equilibrium conditions and self-organization processes at driven out-of-equilibrium conditions.
Order parameter
Order parameter
Surrey et al., Science, 2001
Anomalous Behavior of Supercooled Liquids
Liquid water is often described as anomalous because its behavior frequently departs from that of conventional “simple” liquids. Examples of water anomalies include a temperature of maximum density and increases in thermodynamic response functions (e.g., isothermal compressibility, isobaric heat capacity, etc.) upon isobaric cooling. One influential hypothesis that explains water’s anomalous behavior posits the existence of metastable liquid-liquid phase transition between two (high- and low-density) forms of liquid water (a phenomenon known as “liquid water polymorphism") at deeply supercooled conditions. We are trying to develop an energy landscape perspective of the anomalous thermodynamic and dynamic behavior of water at supercooled conditions.
Machine Learning for Materials Design
Machine learning (ML) is increasingly becoming central to the inverse design of functional materials, providing a predictive framework for engineering systems with desired structural and functional properties. ML techniques enable the discovery of collective variables (such as order parameters) that effectively capture phase transition pathways. When integrated with enhanced sampling methods, these data-driven descriptors enable efficient exploration of the high-dimensional configuration space, making it possible to probe rare events such as nucleation at significantly reduced computational cost. Using ML-based approaches, we are developing rational design strategies to control phase transition pathways and selectively guide the formation of targeted self-assembled structures.
Self-Assembly at Nanoscales
Self-assembly processes play a crucial role in designing many of the multi-scale complex structures in nature. In these processes, solvent-mediated effective inter-particle interactions guide the nanoscale building blocks to spontaneously assemble into the target structure. Therefore, it is crucial to have control of the inter-particle interactions between the self-assembling agents to get the target structure. Soft nanoscale particles are very promising for designing functional materials because the interaction between the building blocks can be easily altered. Using computational modelling, my research group is currently working on to develop rational design and selection criteria for solvent-mediated pair interactions between the nanoscale building blocks that would lead to the desired free-energy landscape, and in turn, the desired structure and properties.
Target Structure
Self-assembling agents
Effective interaction?