Machine Learning (ML) technologies promise to enable chemists to use chemistry reactivity data to predict effective conditions (catalysts, bases, solvents and temperatures, for example) for the synthesis of new, important molecules.  Existing databases are not up to the task of providing data for ML modeling, as the data does not contain information on reaction failure, and is not well-controlled. Pro-active generation of data is an exciting possibility, but to date, chemistry experimentation and analytical technologies have not been up to the task of generating large systematic multi-dimensional data sets that can be subjected to modern ML-protocols.  We have created new ultra-highthroughput experimentation tools that can deliver large data sets that, for the first time, can be modeled to provide high quality predictive models.   These technologies will be described, as well as how we applied ML to study the C-N coupling reaction in great detail. <br/>

Pro-active mapping of chemical reactivity with ultra-highthroughput chemistry

256th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.