Prediction of higher selectivity catalysts by a computer driven workflow and machine learning and computational investigation of boronate esters in the Suzuki-Miyaura cross coupling

Chapter one of this work provides a comprehensive overview of chemoinformatics in enantioselective catalysis. Chapter two is comprised of completely unpublished work detailing the synthesis of a diverse set of bisoxazoline ligands in the planned optimization of an enantioselective aziridination reaction. Although this work was unsuccessful in the optimization of the target reaction, it revealed deficiencies in our computationally-guided workflow. In Chapter three, we address the limitations revealed in chapter two using an algorithmically selected set of BINOLphosphoric acids to simulate an optimization of an enantioselective reaction. In this chapter, we demonstrate the ability to use suboptimal reaction results to predict reaction outcomes for optimal catalysts. In Chapter 4, we transition from chemoinformatics-guided optimization to applied quantum chemistry to elucidate the influence of boronic ester structure on the rate of transmetalation. Here we find that the activation barrier for transmetalation is dependent on the interplay between ground state destabilization of a Pd-O dative interaction and hyperconjugative activation of the Cipso-B σ bond.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Chemistry

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2020

Author dc:creator

• Zahrt, Andrew F

Contributors dc:contributor

• Denmark, Scott E
• Burke, Martin D.
• Peng, Jian
• Pogorelov, Taras V.

• Machine learning
• Computational Chemistry
• Chemoinformatics

Statement dc:rights

• Copyright 2020 Andrew Zahrt

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/108577

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/108577