D3SC: Dynamic Effects in Ordinary Organic Reactions in Solution

在化学系化学结构、动力学和机理-B计划资助的这个项目中，丹尼尔A.德克萨斯A M大学的辛格尔顿将研究反应机制中的“动力学效应”。这些都是实验观察到的反应，既不能预测，也不能解释使用标准模型的机制。 该研究有五个与动态效应相关的目标。 前两个目标将探索新发现的动力学效应形式，一种是由光促进的反应中可能发生的局部分子振动引起的，另一种是由现代统计速率理论的缺陷引起的。第三个目标是应用新发展的动态效应的理解，在溶剂中重新解释一个大的经典化学领域，特别是20世纪世纪碳阳离子化学。在这个广泛的领域，有人认为，旧的争议产生于使用过于简化的机械模型。 第四个目标是应用一种动态效应产生的技术来了解当前化学方法所隐藏的机制。 总体目标是通过将当前理论与机器学习相结合，开发一个全面的动态效应化学理论。总体而言，该项目将在基础和广泛的层面上影响对许多经典化学反应的理解，同时为反应的发展和控制提供新的方法。学生培训是这项研究的一个主要目标，包括将妇女和科学领域代表性不足的群体成员作为研究团队的一部分。 研究生和本科生在单身组进行全面的项目，从而导致不同技能的发展。 这些范围从实验有机化学和精密分析测量，到量子化学，编码，分子动力学计算和机器学习。 Singleton小组还将继续其开发和分发社区计算工具的计划，这些工具有助于研究反应中的动态效应。该研究将特别研究一种间接方法，以实现长期寻求的目标，即在复杂分子中振动促进特定反应。要探索的想法是，不同能量的三重态光敏剂可以为分子中的局部区域提供可控量的能量。该项目还将研究变分过渡态理论中的缺陷，其中实际反应路径偏离基于势能面的预期反应路径，其主要目标是确定受此缺陷影响的反应的其他例子。第三个项目的目标是基于这样的建议，即碳阳离子化学中的历史争议是由于不完全溶剂化结构的未考虑的中间性造成的。这个想法将在一系列突出的经典碳阳离子反应中进行探索。 第四个项目的目标是利用最近开发的动态效应“能量读出”的应用，研究涉及烷氧基自由基的反应中的关键隐藏步骤。这一想法特别适用于非绝热质子耦合电子转移（PCET）反应，其中的机制是无法通过普通的实验和计算方法。 最后的目标是通过使用机器学习将过渡状态划分为动态结果区域来修改过渡状态理论。这包括开发主动学习和提高精度的方法，旨在简化动态效应轨迹的应用，以及为更广泛的化学界提供这些计算工具。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

In this project funded by the Chemical Structure, Dynamics, and Mechanism-B Program of the Chemistry Division, Professor Daniel A. Singleton of Texas A&M University will investigate “dynamic effects” in reaction mechanisms. These are experimental observations in reactions that can neither be predicted nor explained using the standard models for mechanisms. The proposed research has five aims related to dynamic effects. The first two aims would explore newly identified forms of dynamic effects, one resulting from localized molecular vibrations that can occur in reactions promoted by light and another that arises from a flaw in modern statistical rate theories. A third aim is to apply newly developed understanding of dynamic effects in solvent to reinterpret a large classical area of chemistry, specifically 20th century carbocation chemistry. In this broad field, it is suggested that old controversies arose from the use of mechanistic models that were too simplified. A fourth aim is to apply a technology that arises from a dynamic effect to learn about mechanisms that are hidden from current chemical approaches. The overarching aim is to develop a comprehensive chemical theory of dynamic effects by combining current theories with machine-learning. Overall, this project will impact at a fundamental and broad level the understanding of many classic chemical reactions, while providing new approaches to the development and control of reactions. Student training is a major goal of this research including with a good record of including women and members of underrepresented groups in science as part of the research team. Graduate and undergraduate students in the Singleton group undertake comprehensive projects, which lead to the development of diverse skills. These range from experimental organic chemistry and precision analytical measurements, to quantum chemistry, coding, molecular dynamics calculations, and machine learning. The Singleton group will also continue its program of developing and distributing to the community computational tools that aid the study of dynamic effects in reactions. The research will particularly investigate an indirect approach to the long-sought goal of vibrationally promoting specific reactions in complex molecules. The idea to be explored is that triplet photosensitizers of varying energies can provide a controllable amount of energy to localized regions in molecules. The project will also investigate a flaw in variational transition state theory in which the actual reaction path departs from that expected based on the potential energy surface, with the primary goal of identifying additional examples of reactions subject to this flaw. The third project aim is based on the proposal that historical controversies in carbocation chemistry have resulted from the unconsidered intermediacy of incompletely solvated structures. This idea will be explored within a series of prominent classical carbocation reactions. The fourth project aim is to use a recently developed application of a dynamic effect, “energy read-out,” to study key hidden steps in reactions involving alkoxy radicals. This idea is notably applicable to non-adiabatic proton coupled electron transfer (PCET) reactions where mechanisms are inaccessible through ordinary experimental and computational methods. The final aim is to modify transition state theory by partitioning the transition state into regions of dynamical outcome using machine learning. This includes the development of active-learning and precision-enhancing methods designed to ease the application of trajectories to dynamic effects, as well as the provision of these computational tools to the wider chemical community.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.