New conformational sampling and large-scale electronic structure techniques: applications to polypeptide structure, proton transport, and dynamics of silicate melts

纽约大学的Mark Tuckerman教授在理论与计算化学的支持下，对多肽的结构、质子传输和硅酸盐熔体的动力学进行了基于量子力学的研究。这项研究的一个重要组成部分是处理目前限制这种计算效率的一般算法问题。该研究涉及发展动态力技术，允许人们使用由量子力学处理电子自由度决定的力，以半经典或量子力学的方式处理原子和振动自由度。特别强调的是发展的离散变分网格的解决薛定谔方程和数值缩放技术，有利于模拟罕见的事件。本研究的目的还在于局部正交分子轨道的计算。这项工作涉及新的计算算法的发展，这些算法将用于研究燃料电池设计、生物分子模拟和核废料储存中面临的看似不同的技术问题。与这三个主题相关的共同问题是，人们最感兴趣的是理解与电子相互作用有关的不经常发生的过程，以及确定如何控制这些过程的速率。

Professor Mark Tuckerman, of New York University, is supported by Theoretical and Computational Chemistry to perform quantum-mechanical-based investigations on the structure of polypeptides, proton transport and the dynamics of silicate melts. A significant component of this research deals with general algorithmic issues that currently limit the efficiency of such calculations. The research deals with developing on-the-fly force techniques that allow one to treat the atomic and vibrational degrees of freedom semiclassically or quantum-mechanically using forces that are determined from quantum-mechanical treatment of the electronic degrees of freedom. Particular emphasis is on development of descrete-variational grids for solution of Schrodinger's equation and numerical scaling techniques that facilitate the simulation of uncommon events. The effort also aims at the calculation of localized orthogonal molecular orbitals.This work deals with development of new computational algorithms that will be used to study the seemingly disparate technical problems confronted in the design of fuel cells, the simulation of biological molecules and the storage of nuclear waste materials. The common problem associated with these three topics is that one is most interested in understanding processes that, in relation to electronic interaction, occur infrequently and in determining how to control the rate of these processes.