An Asymmetric Route to Xantholipin B, Prediction of Radical Reactions with Machine Learning, and Metal-Catalyzed Carbene Insertion into Aliphatic N−H Bonds
- Author(s): Kadish, Dora
- Advisor(s): Van Vranken, David L
- et al.
Throughout my time at UCI I have been involved in three projects: development of an asymmetric route to xantholipin B, asymmetric insertion of carbene groups into aliphatic N−H bonds, and developing a pedagogical training set for organic radical chemistry for prediction of reaction products and mechanistic pathways. In the Pronin group I optimized conditions for addition of a tri-n-butylstannylzinc reagent to a MOM-protected salicylaldehyde. In the Van Vranken group I synthesized and characterized a chiral guanidine catalyst useful for asymmetric catalysis. I then focused my efforts on the development of a prediction for prediction of step-wise organic reaction mechanisms based on machine learning. In order to improve the identification of electrophilic and nucleophilic atoms, I showed that easily calculated methyl ion affinities correlated with Mayr’s solution phase reactivity parameters, E and N•sN. I then helped to expand a training database of mechanistic radical reaction steps based on the SMIRKS formalism and test the system called Reaction Predictor once the training was complete.