Abstract

Most homogeneous catalysts derived from organic or organometallic moldecules have been discovered through serendipity or trial and error, rather than by design. Computational methods, incorporating both classical simulation and electronic structure theory, however, have become a useful tool for understanding and predicting the roles of such catalysts in chemical reactions. In addition to achieving reactivity, newly developed catalysts must also be competent in selecting for a single product amongst a multitide of potential competing reaction types (chemoselectivity) and alternative positions of reactivity (regioselectivity). Control over 3D-structure and the optical purity of chiral molecules formed by catalytic reactions (stereo- and enantioselectivity) is particularly desirable in modern chemistry and requires a precise understanding of the transition structures in these reactions.1 I will discuss catalysis of some basis ring-closing reactions : computational insights into mechanism have been used to explain unusual reactivity (such as a formally disfavoured 5-endo-trig carbocyclization),2 and also in the design of more atom-efficient asymmetric primary amine catalysts.3 In a transition-metal catalyzed cycloisomizeration [5+2] cyclizations of ynamides, we have discovered how electronic modulation of the organic ligands can be used to influence reaction rate and enantioselectivity, validated in experiments.4 Due to the conformational degrees of freedom in these systems, we use quantum-guided molecular mechanics to parameterize force fields for systems from reference data, and transition structures of interest