Computational Aspects of Potential Energy Surfaces

在这个物理化学项目中，McIver将开发旨在有效评估Born-Oppenheimer势能面的方法。该结果将用于计算多原子速率常数，包括隧道效应，用于使用变分过渡态理论计算一些涉及自由基和双基的反应。这种计算的关键瓶颈是反应路径哈密顿量的评估，这是目前几乎所有使用的多原子速率理论的基础。针对这一瓶颈，我们将采取三条进攻路线：发展一种形式，将反应路径哈密顿坐标（反应路径和横向振动）和耦合常数与势能导数明确地联系起来；利用自洽场波函数加速势能面计算技术的发展并在分布式存储并行计算机上实现了上述方法。计算机代码将在符号代数软件和并行计算机模拟器的帮助下开发，两者都在一个适度配置的工作站上运行。

In this project in the Physical Chemistry Program, McIver will develop methods aimed at efficiently evaluating Born-Oppenheimer potential energy surfaces. The results will be used in computing polyatomic rate constants, including tunnelling effects, for a number of reactions involving radicals and biradicals using Variational Transition State Theory. The critical bottleneck in such calculations is the evaluation of the Reaction Path Hamiltonian which underlies virtually all currently used polyatomic rate theories. Three lines of attack on this bottleneck will be pursued: development of a formalism which explicitly relates the Reaction Path Hamiltonian coordinates (the reaction path and transverse vibration) and coupling constants to potential energy derivatives; development of techniques for speeding up the calculation of potential energy surfaces using self consistent field wavefunctions; and implementation of the methods on a distributed memory parallel computer. The computer codes will be developed with the aid of symbolic algebra software and a parallel computer simulator, both of which operate on a modestly configured workstation.