Author: Kaid C. Harper, Matthew S. Sigman*
Affiliation: Department of Chemistry, University of Utah, 315 South 1400 East Rm. 2020 Salt Lake City, Utah 84112-0850
In chemistry today, there is often a need for the synthesis of only one enantiomer in a given reaction, especially in the field of pharmaceuticals. Asymmetric synthetic strategies often employ a catalyst that consists of a metal with a ligand scaffold that transmits chiral information to the products, subsequently producing a high enantiomeric ratio (e.r.). Because the difference in energy between the two diastereomeric transition states that lead to separate enantiomers can be very small, electronic and steric effects play a magnified role in asymmetric catalysis, for example, by either lowering the transition state energy of the desired enantiomer through electronic stabilization or by raising the transition state energy of the undesired enantiomer through steric repulsion. Though the design of asymmetric catalysts often takes into consideration both of these effects independently, they are generally developed on a trial-and-error basis with little theoretical guidance. Unfortunately, this approach is time and resource consuming and an increased emphasis has been placed on developing methods to rationally design chiral catalysts.
The authors of this article have done exactly this, employing computer models to fit data points that correlate electronic and steric effects of a given catalyst using classical physical organic parameters that in the past have been considered independently of one another. In this method, electronic effects (E) are described by Hammet principles that rely on the theory that the relative rates of reactions can be predicted based on the linear relationship between the free energy barrier of a reaction and changes in substituents on the reactants. Steric effects (S) are described by Taft-Charton parameters that relate Van Der Waals radii of the substituents to a specific value. The optimization of an enantioselective Nozaki-Hiyama-Kishi propargylation of ketones is used to test their method, in part because of the many synthetic applications of homopropargylic tertiary alcohols, and also because an enantioselective method using commercially available reagents had not been previously reported.
Their approach combined experimental design principles with the systematic synthesis of a ligand library containing ligands with varying E and S values. By creating mathematical surface models, a ligand with optimal E and S character can be predicted to maximize ΔΔG‡ without previous knowledge of the mechanism. The data they collected was fit to a third-order polynomial function shown below, with an impressive correlation coefficient of r2 = 0.96.
The surface generated predicted an optimal ligand (shown in the figure) that performed very well over a broad scope of aryl and heteroaromatic methyl ketones giving e.r.s in the 90s with good yields (60-86%). These strong results show that this method will be a great tool for predicting the best balance of electronic and steric factors for optimally selective catalysts while saving time and resources.