(6gb) Understanding Molecular Mechanisms of Rare Events: Applications in Nanofabrication in Semiconductor Industry, Novel Drug Delivery Systems and Fuel Cells

At molecular level, rare events are events which occur at timescales of a few nanoseconds to several days. These events underlie many observed physical phenomena, such as phase transitions, conformational changes in proteins and polymers, chemical and biological reactions and self-assembly.  Rare events often have a rate limiting step, which lies at the maxima of the free energy along the reaction coordinate pathways. Since traditional molecular simulation methods are ineffective in sampling such events, special techniques have been developed to sample them, such as forward flux sampling, variants of string method and transition path sampling. Deciphering molecular mechanisms behind rare events is fundamental to our understanding of physical systems of interest and useful for many technological innovations. In the last four decades, semiconductor industry has seen a relentless scaling down of transistor dimensions in accordance with the Moore’s law. The state of the art semiconductor technology is operating at critical dimensions of tens of nanometers. The traditional scaling of transistor dimensions, often called Dennard scaling, is no longer applicable at present length scales, and hence focus has moved to innovations in novel materials and new fabrication methodologies. One approach is to use a high K dielectric material in thin film transistors. Hafnium oxide based inorganic dielectric materials have many problems, such as poor interface properties with silicon causing development of interface traps, which affect reliability of devices. Molecular modeling methods can be utilized to design optimal polymeric substances which are capable of forming a good interface layer with silicon. High-K polymeric materials, such as poly(vinylidenefluride-co-trifluoroethylene) and poly vinyl phenol with varying molecular weight, grafting density and co-polymeric mixtures would be materials of interest for this study.  Long relaxation times of these polymers as well as their conformational changes under electric field can be studied as rare events in molecular simulations. Another innovation which has improved performance of semiconductor devices is the introduction of strain in silicon to increase mobility of charge carriers. One way to induce strain is by first depositing thin films and then either curing them or by providing a free surface to allow them to expand or contract. Molecular modeling of structural modifications in thin films via rare event sampling techniques can provide useful insights on the role of chemistry of the film and its interaction with the underlying silicon crystal, which will help in better design of this process. Hydrogels are cross-linked network of hydrophilic polymers. Hydrogels can undergo gel-sol transition upon changes in environmental conditions, such as pH, temperature or light. Using environmental sensitive hydrogels as site specific drug delivery systems is a promising idea. The important parameters of investigation would be the sensitivity of sol-gel transition to stimuli, and the kinetics and thermodynamics of the transition. These investigations can most effectively be undertaken by simulating the sol-gel transition under varying environmental conditions using molecular simulations. Transport properties of membranes are functions of the structure of their polymer network. For fuel cell applications, it is desirable to have high proton conductivity with reduced water cross-over. Molecular modeling of proton transport as a function of the structure of the membranes would help develop more selective membranes with high permeability for protons for fuel cell applications. In my post-doctoral work at Princeton University, I applied forward flux sampling in molecular dynamics simulation of an all-atom water model to study kinetics of evaporation and condensation of water in hydrophobic confinement at ambient conditions, thereby simulating time-scales in order of few hours. Using transition path sampling, I determined transition state ensemble associated with liquid to vapor transition of water in hydrophobic confinement. In my doctoral work at Columbia University, I studied conformational changes in proteins upon adsorption to hydrophobic surfaces. In order to eliminate kinetic traps while sampling different conformational states of proteins, I did Monte Carlo simulation at different temperatures and surface hydrophobicity to study different regions of the phase space. The entire phase space is generated by using methods such as simulated annealing, weighted histograms and parallel tempering. In order to study kinetics of a system which can transition between two states, the phase space can be first sampled by using equilibrium sampling techniques such as parallel tempering, simulated annealing and weighted histograms. Once different minima in the phase space have been identified, rare event sampling can be used to study the kinetics of transition from one minimum to the other.