Unbiased Molecular Discovery: Laboratory and Machine Learning Approaches to Expanding Our View of Interstellar Chemistry

As the sensitivity and spectral grasp of our radio telescopes continues to expand, so too does our window on the shocking degree of chemical complexity shaping and being shaped by the process of forming stars and planets. Last year we reported the first detections of individual polycyclic aromatic hydrocarbon molecules in space. These species are thought to be the reservoir of a vast fraction of interstellar organic carbon, but our understanding of the chemistry leading to their formation is shockingly limited, largely by a dearth of detections of both these kinds of molecules and their probably reaction intermediates. Here, I will describe how our collaboration -- GOTHAM -- is tackling this challenge on two fronts. In the laboratory, our team members are using Microwave Spectral Taxonomy to perform unbiased reaction screening studies to identify new molecules of potential interest and provide the spectra necessary to study them observationally. As well, our team has developed novel machine learning approaches to reproducing and, critically, predicting, the chemical inventories of molecularly rich interstellar sources with astonishing accuracy.