Chemical Cartography via High-Throughput Experimentation: Predictive Models, Catalyst Development, and New Synthetic Methodology

点击翻译按钮获取中文摘要

Organic synthesis is among the most impactful scientific developments in history, dramatically improving quality of life via breakthroughs in medicine, agriculture, and materials. Despite these advances and more than a century of research, in most labs the practice of organic synthesis is remarkably unchanged from how it was done in the early 1900s. Individual chemical reactions are optimized through an iterative and often trial-and-error approach using single experiments carried out in flasks. While this has worked well in the past, we are now at a point where continued progress in the field requires new, more efficient techniques and tools to enable a deeper understanding of chemical reactivity. The theme of the Leitch group program is "the exploration of uncharted chemical space." This means finding new points on the map (novel chemical structures), and studying the paths between these points (chemical reactivity). Specifically, we will address a centrally important but still unsolved problem in organic chemistry: how can one predict chemical reactivity in a quantitative manner, and use these predictions to develop new and more efficient chemical syntheses? My group will tackle this problem by combining fundamental physical chemistry principles with modern high-throughput experimental methods and data analytics. We will use this approach to generate quantitative mechanistic models - i.e. maps of chemical reactivity - for key chemical reactions currently used for pharmaceutical synthesis, and to develop scalable syntheses of novel three-dimensional carbon frameworks that are at the forefront of modern drug discovery research. Critical to this endeavour is the simultaneous measurement of hundreds-to-thousands of chemical reaction rates and activation energies using high-throughput experimentation. Combining these values with computed molecular parameters for each chemical species will generate large, reliable, and consistent data sets. The size and mechanistic foundation of these data sets will be a distinct advantage in building meaningful quantitative models via algorithm-driven statistical analysis. These models will allow us to predict the outcome of a chemical reaction under a variety of hypothetical conditions, leading to a deeper and more holistic understanding of the factors that control chemical reactivity. The potential impact of this research in both academic and industrial contexts is substantial. The ability to predict the outcome of a given reaction will save countless person-hours in the pursuit of new therapeutics, agrochemicals, and advanced materials. Being able to quantitatively map how chemical structure affects reactivity will enable the discovery of new and more efficient syntheses in a rational manner. Finally, our reactivity maps will be powerful data sets on which to build predictive artificial intelligence systems for chemical synthesis design; this facet is one of the ultimate goals of this program.