Computational study of the atomistic kinetics and structure of complex materials

点击翻译按钮获取中文摘要

This research programme focuses on the development and the application of advanced computational approaches for the study of structural and dynamical properties of complex materials — alloys, disordered systems, defective materials, and proteins. While they appear very diverse, all these systems share a similar problematic : their configurational space is very rich and difficult to explore using standard methods such as molecular dynamics and understanding the atomistic details of their evolution over long time scales (i.e., beyond a microsecond or so) is crucial to the proper characterization of their properties. Achieving these goals on computer, even with access to state-of-the-art computing facilities, remains a challenge and requires new and, often, specific approaches. Over the years, I have developed many algorithms to bridge the gap between computer and experimental time scales that have allowed my group to remain at the forefront of accelerated methods and their application to complex materials. Over the next five years, I propose to concentrate a good deal of my efforts to apply and further optimize the kinetic Activation-Relaxation Technique (k-ART), the only kinetic Monte-Carlo algorithm able today to simulate, without any restriction, the kinetics of glasses, interfaces and alloys over time scales of a second or more. With such a method, I have access to vast families of problems that could not be addressed numerically until now such as the evolution of low-temperature glass and disordered systems, the formation of nanostructures in complex environment (e.g., silicon nanocrystals in silica) and the diffusion of defects and impurities in materials (e.g., self-defects, carbon and helium atoms in iron). I will also continue development on the biophysics front, focusing in large part on further characterizing the aggregation process of amyloid proteins in order to identify the neurotoxic oligomer species. Here, the challenge is also one of time scale and sampling and we have shown that simplified potential, such as the coarse grained OPEP forcefield, could help us push these limits further. However, coarse-grained forcefields also come with limitations and part of this programme will focus on the development of a new all-atom implicit solvent potential that will sit between OPEP and the more realistic descriptions such as CHARMM and AMBER, giving us more precision and remain as light as possible. With this forcefield, we will continue our study of amyloid proteins using a multiple approaches — large assembly of short peptides, and small assembly of full-length with and without membranes — focusing on the detailed comparison between various sequences to extract qualitative results less likely to be finely dependent on the forcefield specifics. Beyond these applications, we will make significant efforts to distribute the various unique tools that we are developing as those are already attracting a lot interest around the world. Maintaining and distributing these codes is time and resource consuming but it is an essential part of method development today. Overall, this research programme proposes a good equilibrium between methodology and physics to ensure that my group remains at the forefront of methods development and continues to deepen our understanding of complex materials.