Minimal Theoretical Description of Magnesium Halogen Exchanges

Although Grignard exchange reactions have been known for about a century, little beyond heuristics guides the prediction of when and where an exchange will occur. Additionally, little is understood from a modeling perspective about why the heuristics work or when they might fail. We develop a simple quantum mechanical model for predicting when and where an exchange will occur. We show that the solvent-free model successfully predicts for reactions regioselectively using only thermodynamic end points. After validating the model, we employ it to explain some of the heuristic understandings, such as preference for exchanging with halogens lower on the periodic table and facilitation of the exchange with lithium chloride. Preference for bromine over chlorine appears to be related to both kinetic barriers and thermodynamic considerations, and addition of lithium chloride to the exchange mixture appears to act almost exclusively in a catalytic role. We then use the model to predict materials that should allow ready exchange onto any carbon, including tertiary halide carbons. We find that removable silicon-linked directing ethers attached to the molecule of interest and, potentially more versatilely, silyl Grignard reagents could serve in this role.尽管格氏交换反应（Grignard exchange reactions）已被发现近百年，但目前除经验性规则外，几乎无其他依据可用于预测该反应发生的时机与位点。此外，从建模视角而言，学界对这些经验规则为何成立、或在何种情境下会失效的理解仍极为有限。我们构建了一款简易量子力学模型，用于预测格氏交换反应发生的时机与位点。研究证实，仅依托热力学终点参数，该无溶剂模型即可成功实现反应的区域选择性预测。在验证该模型的有效性后，我们利用其阐释了部分经验规则的内在机理，例如对元素周期表中位置更靠下的卤素的交换偏好，以及氯化锂对交换反应的促进作用。相较于氯，溴的交换偏好似乎同时与动力学势垒和热力学因素相关；而向反应体系中添加氯化锂，其作用几乎完全体现为催化功能。随后，我们借助该模型预测了可使任意碳原子（包括叔卤代碳原子）上顺利发生格氏交换反应的底物材料。我们发现，连接于目标分子的可脱除硅连接型导向醚，以及通用性更强的硅基格氏试剂（silyl Grignard reagents），均可承担这一功能。