Doctor of Philosophy in Chemistry (PhD)
Statistical Models and Machine Learning for the Design and Optimization of Novel Catalysts
Theses completed in 2010 or later are listed below. Please note that there is a 6-12 month delay to add the latest theses.
Chiral phosphoric acid (CPA) catalysts have witnessed rapid development recently. They have demonstrated remarkable applicability in catalyzing various reactions in part due to their modular chiral environment and acidity through substituent tuning. In the pursuit of understanding their catalytic mechanisms for optimizing reaction conditions and designing improved catalysts, chemists have applied a range of techniques to understand how these catalysts operate. Chiral phosphates, a structurally related but much less understood catalyst system have also been shown to catalyze many transformations in high levels of enantioselectivity. However, the lack of mechanistic understanding makes it difficult to develop reactions relying on these catalysts more generally and devise novel transformations.With this backdrop, the goal of my work is to first develop a comprehensive understanding of how chiral phosphates impart high levels of enantioselectivity for diverse reactions. Then in turn use this information to design reactions and catalysts for higher enantioselectivity in asymmetric catalysis. More specifically, I studied the chiral phosphate-catalyzed aza-Friedel-Crafts reaction using computations. The results showed that in the lowest energy transition states (TS) two hydrogen bonds, one from the iminium (C–H) and the other from indole (N–H), interact with the catalyst oxygens. Expanded exploration of this mode of stereoinduction to other similar reactions involving cationic intermediates show many reactions proceed in this way. Critically, the strength of the C– H contact with the phosphate can impact the level of enantioselectivity afforded. With this information, I set out to develop a complementary reaction to achieve cyclic N,S-acetals in higher enantioselectivity with designer protocols that rely on significantly stronger H–bonding interactions with the catalyst. Although, these efforts did not provide high levels of enantioselectivity for a variety of reasons the reaction strategies should provide a useful starting point for further optimization. Finally, one of the potential issues with the new protocols could be a lack of acidity, meaning background reactions are competitive. Given this and with the communities’ desire to continuously develop Brønsted acid catalysts with greater levels of acidity, I attempted to design and test a new biphenyl catalyst embedded with unique, chiral electron withdrawing substituents.
The scope of chiral phosphate catalysis is rapidly expanding and currently ranges from the electrophilic activation of allenamides to the hydrogenation of enals. Despite the significance of such transformations for the stereocontrolled synthesis of a wide variety of organic compounds, the preferred pathway and reasons for selectivity remain unclear, making it challenging to develop new reactions. Here, I attempt to address this issue by using transition state calculations that provide important 3-dimensional pictures to allow the analysis of several chiral phosphate catalyzed enantioselective transformations. This group of seemingly unrelated reactions often occurs through a single mechanism involving two hydrogen-bonding contacts from the iminium intermediate and nucleophile to the catalyst. I explain the various molecular features that affect enantioselectivity allowing the development of stereochemical models. As noted throughout, my coworkers Professor Jolene Reid and Mr. Jianyu Zhai were able to develop and apply statistical models which allow for precise predictions of enantioselectivity. These quantitative models also provide some mechanistic insight that complements my calculations. This modeling approach that uses a suite of computational techniques should be generally applicable to other catalytic systems that are often used in asymmetric synthesis.