Computational Chemistry Studies of Organometallic Energy Landscape

Computational chemistry is becoming a widely used tool to investigate the kinetics and thermodynamics of chemical transformations. These investigations are often heavily guided by experiment and require significant mechanistic insight prior to meaningful model development. Recent advances in reaction path finding and automated potential energy surface assessment have enabled faster and easier exploration of complex chemical mechanisms. In combination with mechanistic information, structure energy correspondence provides information which describes how a particular reaction mechanism energetically varies as structure is modulated. Together, the relevant reaction pathways and the structure energy relationships describe the reaction landscape for a given class of reactivity.;Chapter 1 introduces the core chemical concepts needed to understand reaction landscapes. The tools and information needed to perform detailed mechanistic exploration via computation are presented and competing methods are summarized. Further discussion of reaction path finding tools is provided through an example involving the reactivity of ammonia borane and carbon dioxide. A discussion of the characteristics which connect potential energy surfaces to quantitative structure activity relationships is used to conclude this chapter.;Chapter 2 details the application of an automated reaction path finding tool for the investigation of intuitive and non-intuitive pathways for C(sp3)-N reductive elimination from palladium(IV). This work demonstrates that detailed computational studies using automated reaction path investigation can be used to assess unexpected reaction pathways. These simulations predicted the relative reaction rates with various sulfonamides through consideration of both intuitive and non-intuitive reaction mechanisms. Overall, this chapter demonstrates that combinations of experimental studies and computational tools can provide fundamental mechanistic insights into complex organometallic reaction pathways. This work begins to explore relevant molecular features which appear to trend well with the experimentally observed reactivity.;Chapter 3 continues the development of molecular feature based investigation. This chapter was inspired by the possibility of using computational investigations of complex organometallic reaction landscapes to describe structure energy correspondence. This section discusses the development of a thermodynamic landscape to investigate CO2 reduction from cobalt bis(diphosphine) complexes. The construction of a dataset of Co(L)(L')H2 type complexes from set of commercially available of bis(diphosphines) covering a thermodynamic landscape of over 50 orders of magnitude acidity and hydricity is discussed. These data suggest that relationships between common steric and electronic molecular features are poorly correlated with catalyst thermodynamics. However, a strong correlation between the thermodynamics and Co---H NLMO energy is observed. The landscape provides a clear example of careful electronic balance required for catalytic relevance. The best catalyst identified for future experimental investigations was Co(dCype)H, which is expected to be more acidic and hydridic than previously reported Co(dmpe)2H.;While there is still significant work remaining in the development of robust and automated computational chemistry tools, this work outlines some potential applications and details the relevant findings. The final chapter discusses the current limitations and challenges associated with computational reaction discovery. Particular attention is paid to the development of reasonable organometallic computational models for use in reaction landscape investigation.