Ensemble-averaged ab initio magnetic resonance and double resonance spectra and line shapes in heterogenous condensed phase systems.

缺乏长程有序的固体和液体中微观结构和分子运动的表征仍然是一个具有挑战性的课题，特别是对于具有许多组分的异质介质。研究这类系统最有力的技术之一是核磁共振（NMR）和电子磁共振（EPR）光谱。为了理解磁共振波谱中的信息，人们需要一种理论将波谱与潜在的分子结构和动力学联系起来。然而，光谱通常是非常复杂的，即使有一个理论在手，它在如此复杂的情况下的确切应用是一项艰巨的任务，远远超出了目前的计算能力。在这种情况下，计算机模拟是非常有价值的。在这个项目中，我提出了沿着这些路线的几个发展，即应用分子模拟方法来理解磁共振和磁双共振谱。在后一种情况下，实验同时涉及到原子核和电子的磁性特征。具体来说，在我们之前工作的基础上，我们的目标是将几种理论技术结合在一起，开发经过仔细测试的方法，帮助解释一些最强大的磁共振技术的结果。我们想要关注的第一项技术是“ENDOR”，这是一种双磁共振技术，能够提供关于分子探针周围异质介质结构的详细信息。首先，我们将解决有关探针结构、凝聚相和光谱之间精确关系的一些基本问题。我们的长期目标是开发一个经过仔细测试的工具箱，从虚拟分子系统中计算ENDOR光谱，并迭代修改分子系统，直到与实验达到最佳匹配。第二种技术是“动态核极化”或“DNP”，也是与ENDOR密切相关的双共振技术。在最近的一项令人惊讶的发现中，人们发现一种特殊风味的DNP在绝缘固体中起作用，这与传统的观点不一致。我们最近处理了这个令人困惑的发现，并成功地解释了它。在这个项目中，我们希望进一步发展我们的模型，并进一步推动可能指导实验发展的新预测。我们要研究的磁共振的最后一个方面是一个更基本的方面。我们希望在凝聚相中原子核之间的间接（电子介导的）磁耦合与氢键的特征、几何形状和共价之间建立定量联系。

The characterization of microscopic structure and molecular motion in solids and liquids that lack long-range order is still a challenging topic, especially for heterogenous media with many components. One of the most powerful techniques to study such systems is magnetic resonance (MR) spectroscopy, both on nuclei (NMR) and electrons (EPR).In order to understand the information in a magnetic resonance spectrum, one needs a theory to connect the spectrum to the underlying molecular structure and dynamics. However, the spectra are usually very complicated, and even with a theory in hand, its exact application for such complicated scenarios is a formidable task that is well beyond present day computational capabilities. In such scenarios computer simulations are very valuable. In this project, I am proposing several developments along these lines, that is the application of molecular simulation methods to understand magnetic resonance and magnetic double resonance spectra. In the latter case the experiment involves simultaneously the magnetic signature of atomic nuclei and of electrons.Specifically, building on our previous work, what we aim to achieve is to combine several theoretical techniques together, developing carefully-tested methods that can aid in interpreting the results of some of the most powerful magnetic resonance techniques.The first technique that we would like to focus on is "ENDOR", a magnetic double resonance technique that has the capacity to provide detailed knowledge about the structure of heterogenous media around a molecular probe. First we will address some fundamental questions regarding the precise relation between the structure of the probe, the condensed phase, and the spectrum. Our long-term aim here is to develop a carefully tested toolbox that calculates the ENDOR spectrum from a virtual molecular system, and iteratively modifies the molecular system until the best match with experiment is achieved.The second technique is "dynamic nuclear polarization" or "DNP", again a double resonance technique that is closely related to ENDOR. In a surprising recent finding, one particular flavor of DNP was discovered to be operative in insulating solids, which disagreed with more traditional wisdom. We have recently addressed this perplexing finding and successfully explained it. In this project we would like to further develop our model, and push further towards new predictions that can possibly guide the experimental developments.The final aspect of magnetic resonance that we would like to investigate is a more basic one. We would like to establish a quantitative link between the indirect (electron-mediated) magnetic coupling between atomic nuclei in condensed phases, and the characteristics of the hydrogen bond, its geometry, and its covalency.