Minimal Theoretical Description of Magnesium Halogen Exchanges

Abstract

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.