Computational Models for New Developments in Radical Chemistry

项目摘要 这个项目的目标是开发新的理解和预测模型，以形成，反应，和 有机自由基和双自由基中间体的选择性。三重双自由基可以经历各种不同的 在Singlet基本状态下不可访问的转换。然而，开发高效的光催化 三重态能量转移过程，特别是以对映体选择性的方式，仍然是一个持久的目标。我们会 说明可以利用计算方法来开发底物和增感剂的一般原理 设计利用三态反应性。我们将使用计算来定向机制引导的发现 三重态中的未知反应性，如均裂芳香取代和手性Lewis的设计 酸促进不对称光催化环化反应。我们还将制定一种定性和定量的 了解控制自由基均裂取代反应速率和位置选择性的因素， 如氢和卤素原子转移反应。这些概念性的见解将被用来合理化 位选择性C(SP3)-H新型自由基试剂的实验观察与开发 氯化作用。在新的物理-有机参数的辅助下，可以使用定量模型的开发 以加速这一进程。基于机械理解的机器学习模型将提供新的 将底物和试剂参数化的工具，以加速反应发现和优化。我们将聘用 这一策略用于预测P450氧化小分子的位置选择性并建立通用工作流程 以预测代谢降解途径。

Project Abstract The goal of this project is to develop new understanding and predictive models for the formation, reactivity, and selectivity of organic radical and diradical intermediates. Triplet diradicals can undergo a variety of transformations that are not accessible in the singlet ground state. However, developing efficient photocatalytic triplet energy transfer processes, particularly in an enantioselective fashion, remains an enduring goal. We will show that computational approaches can be leveraged to develop general principles for substrate and sensitizer design to harness triplet-state reactivity. We will use computation to target the mechanism-guided discovery of unexplored reactivity in the triplet state, such as homolytic aromatic substitution, and the design of chiral Lewis acids to promote asymmetric photocatalytic cyclizations. We will also develop a qualitative and quantitative understanding of the factors controlling the reaction rates and site selectivities of radical homolytic substitution, such as hydrogen and halogen atom transfer reactions. These conceptual insights will be used to rationalize experimental observations and underpin the development of new radical reagents for site-selective C(sp3)-H chlorination. The development of quantitative models, aided by new physical-organic parameters, can be used to accelerate this process. Machine learning models grounded in mechanistic understanding will provide new tools to parametrize substrates and reagents to accelerate reaction discovery and optimization. We will employ this strategy to predict the site-selectivity of P450 oxidation small molecules and to establish general workflows to predict the metabolic degradation pathways.