The Impact of Computational Uncertainties on the Enantioselectivity Predictions: A Microkinetic Modeling of Ketone Transfer Hydrogenation with a Noyori‐type Mn‐diamine Catalyst

Abstract Selectivity control is one of the most important functions of a catalyst. In asymmetric catalysis the enantiomeric excess (e.e.) is a property of major interest, with a lot of effort dedicated to developing the most enantioselective catalyst, understanding the origin of selectivity, and predicting stereoselectivity. Herein, we investigate the relationship between predicted selectivity and the uncertainties in the computed energetics of the catalytic reaction mechanism obtained by DFT calculations in a case study of catalytic asymmetric transfer hydrogenation (ATH) of ketones with an Mn‐diamine catalyst. Data obtained from our analysis of DFT data by microkinetic modeling is compared to results from experiment. We discuss the limitations of the conventional reductionist approach of e.e. estimation from assessing the enantiodetermining steps only. Our analysis shows that the energetics of other reaction steps in the reaction mechanism have a substantial impact on the predicted reaction selectivity. The uncertainty of DFT calculations within the commonly accepted energy ranges of chemical accuracy may reverse the predicted e.e. with the non‐enantiodetermining steps contributing to e.e. deviations of up to 25 %.


S1. Gibbs Free Energies to Rate Constants
To determine rate constants in the microkinetic model, the Eyring equation is considered to calculate the rate constants:

S3. Kinetic trajectories
The kinetic trajectories of enantiomeric excess to conversion were compared to the experimental trajectory mentioned above to obtain the enantiomeric excess at 70% conversion and the RMSD between the data points. A few examples of the trajectories are illustrated below. Figure S1. Trajectories of enantiomeric excess developing as a function of conversion of the experimental data (red) and various microkinetic model runs.

S4. Elementray reactions
Below, the reaction network and the corresponding elementary reactions for the ketone reduction pathway are summarized. The Gibbs free energies profile are presented for both the NH1 and NH2 channel.
S5 Figure S2: Elementary reaction steps Figure S3: Free energy diagram of the reduction of acetophenone proceeding through NH1 channel S6 Figure S4: Free energy diagram of the reduction of acetophenone proceeding through NH2 channel [ [acetophenone]

S5. Degree of rate control
The degree of rate control grouped in the general reaction steps is summarized in the table below: Reaction step DRC TS 1-2 0.88 TS 3-4 0 TS 4-5 0.12 TS 6-1 0

S6. Comparison NH1 and NH2 channel
The proposed reaction networks implies that the reaction can follow several channels. We compared the rate of the reaction proceeding solely through the NH1 and the NH2 channel. Our analysis reveals that the reduction reaction through the NH2 channel proceeds much faster than through the NH1 channel With the NH1 to NH2 ratio of approximately 0.01. The enantiomeric excess at 70% conversion is 41% S-enantiomer when proceeding through the NH1-channel and 80% R-enantiomer when proceeding through the NH2-channel. The reaction proceeding through the NH1 channel can therefore contribute to the deterioration of the enantiomeric excess of Rphenyl ethanol.
S9 Figure S5. Rate of phenylethanol production as a function of conversion proceeding separately through NH1 channel (black) or NH2 channel (red).

S7. Comparison Backward to Forward Reaction Contribution (R and S)
Here reaction rates for the formation of R-phenylethanol and S-phenylethanol are compared. The illustrations below account for the reaction rates leading to the formation of R-and S-enantiomer as a function of conversion. The change of the netto reaction reate for the R-enantiomer is displayed to account for the changes that occur as the reaction proceeds. 3. An ab initio MD run at 333.15K at NVT ensemble was carried out for 5000 steps in cp2k 6.1 software [3] with the quickstep module and orbital transformation for faster convergence. in the initial 500 steps velocity softerning was applied. The DZV-GTH-PADE basis set was applied to Mn, TZVP-GTH to all other atoms. All calculations were spin-polarized and PBE-D3(BJ) [4] correction was applied. 4. The 50 lowest conformers obtained in step 500 to 3000 were taken from the MD trajectory, optimized with constraints in DFT, and the structure resulting in the highest RMSD from the initial transition state structure was fully optimized in accordance to the other reported TS structures.
This search resulted in a 5 kJ mol -1 difference between the initial and the transition state found from the MD run. Energetics and xyz coordinates are reported below.