Spectra, Dynamics, and Intermolecular Forces in Small Molecules and Clusters

了解分子间的作用力或分子内部和分子之间的相互作用能，对于理解和预测从简单的二聚体到稀气体到散装液体或固体等系统的性质是必不可少的。开发和应用方法来确定和表示这种相互作用，并利用这些信息来研究简单分子，无论是孤立的，簇或凝聚相环境，是本提案的中心主题。1)双原子分子建模，数据分析和计算方法。该项目以数据分析的“直接潜在拟合”方法的开发和应用为中心，其中，实验光谱与基于灵活解析模型势能函数的精确量子模拟结果的比较，用于优化定义模型势能函数和原子质量相关的Born-Oppenheimer击穿校正函数的参数，以使不同同位素的数据相互一致。这种方法最早是在我们（1974）对三维原子硅藻范德华分子的处理中引入的，在过去的20年里，它在硅藻中的应用已经在世界各地的许多实验室中得到了发展。与传统的“参数拟合”分析相比，它提供了更基本的物理信息，并产生了更好的外推和内插预测。然而，所使用的一些潜在形式的局限性和方法的复杂性阻碍了它的普遍采用。我们的工作包括开发一种更好的求解径向薛定谔方程的方法，这种方法可以像定位和处理位于势阱深处的能级一样容易地定位和处理离解极近的能级。它还包括开发和测试改进的势函数模型，这些模型包含了正确的理论上已知的远程行为。计划中的工作包括对该模型的扩展，使其能够更好地处理非标准的“潜在货架”、双最小值或屏障行为。2)分子簇和矩阵的动力学和振动位移模拟。对分子（“范德华”）团簇的物理性质和行为最具鉴别性的实验探测是在团簇内或团簇上的发色团的光谱。然而，为了充分利用实验工作，需要定量了解发色团分子的光谱特性如何受到其周围环境的影响。理解和预测被困在固体内部或表面的分子的振动频移是一个类似的问题，因为两者都是由相同类型的微观分子间相互作用和频移机制控制的。对这种变化的定量理解将有助于探测材料表面和内部的局部秩序，确定有限分子团簇的相行为（类液体或类固体），并检查团簇型系统中的溶剂化行为。我们之前的工作包括开发了一个成功的模型，用于定量预测簇环境中某种类型的发色团的红外光谱的分裂和移动；结果表明，这种光谱可以为簇的大小、结构和相行为提供明显的特征。我们现在正在应用路径积分量子蒙特卡罗方法来模拟分子团簇中红外发色团的性质和光谱特征，并发现它们提供了对相互作用的微妙之处和动力学行为的本质以及可能存在的超流动性的尖锐测试。

Knowledge of the intermolecular forces or interaction energies within and between molecules is essential for understanding and making predictions about properties of systems ranging from simple dimers to dilute gases to bulk liquids or to solids. Developing and applying methods for determining and representing such interactions, and utilizing such information to study simple molecules, both isolated and in clusters or condensed phase environments, is the central theme of this proposal.1) Diatomic Molecule Modelling, Data Analyses, and Computational Methods.This project is centred around the development and application of the 'direct potential fit' method of data analysis, in which a comparison of experimental spectra with the results of an exact quantum simulation based on a flexible analytic model potential energy function is used to optimize parameters defining both the model potential function and the atomic mass-dependent Born-Oppenheimer breakdown correction functions required to make the data for different isotopologues consistent with one another. This approach was first introduced in our (1974) treatment of three-dimensional atom-diatom Van der Waals molecules, and its applications to diatomics have been developed over the past 20 years in a number of laboratories around the world. It provides more fundamental physical information, and yields much better extrapolation and interpolation predictions than do conventional 'parameter-fit' analyses. However, the limitations of some of the potential forms used and the complexity of the method have inhibited its general adoption. Our work has included the development of a better way of solving the radial Schroedinger equation that allows levels lying extremely close to dissociation to be located and treated just as easily as levels lying deep in a potential well. It has also included the development and testing of improved potential function models that incorporate the correct theoretically-known long-range behaviour. Planned work includes an extension of this model to allow it to better deal with non-standard 'potential shelf', double-minimum or barrier behaviour. 2) Simulations of Dynamics and Vibrational Shifts in Molecular Clusters and Matrices.The most discriminating experimental probe of the physical properties and behaviours of molecular ('Van der Waals') clusters are the spectra of chromophores seeded into or onto the cluster. However, to exploit fully the experimental work requires a quantitative understanding of how the spectroscopic properties of the chromophore molecule are affected by its surroundings. Understanding and predicting the vibrational frequency shifts of molecules trapped within or on the surface of solids is a similar problem, since both are governed by the same types of microscopic intermolecular interaction and frequency-shift mechanisms. Quantitative understanding of such shifts would facilitate probing local order in and on the surface of materials, determining the phase behaviour (liquid-like vs. solid-like) of finite molecular clusters, and examining solvation behaviours in cluster-type systems. Our previous work included the development of a successful model for quantitative prediction of the splitting and shifting of IR spectra of a certain type of chromophore in a cluster environment; it showed that such spectra could provide distinct signatures for cluster size, structure and phase behaviour. We are now applying path-integral quantum Monte Carlo methods to simulate the properties and spectroscopic signatures of infrared chromophores in molecular clusters, and have found that they provide sharp tests of subtleties of the interactions and the nature of the dynamical behaviour and possible presence of superfluidity.