Similarity-based spectroscopic fingerprints map reactivity for autonomous chemistry
A new paper from the Cronin group in Angewandte Chemie International Edition introduces a chemically agnostic way to monitor and classify reactions using only similarity measures applied directly to spectroscopic data. By tracking simple similarity trajectories (such as Jaccard indices and cross-correlations) on in situ NMR, IR, UV/Vis and EPR measurements collected on the Chemputer platform, the team can follow reaction progress, detect endpoints and distinguish different kinetic regimes, including first-order decay, autocatalysis and oscillatory behaviour, without any prior mechanistic knowledge or peak assignment. The approach is demonstrated across a wide range of systems, from lophine formation and azide synthesis to the Belousov–Zhabotinsky reaction and a fuel-driven [2]rotaxane molecular switch, providing a general framework for turning high-dimensional spectroscopic data into quantitative “reactivity fingerprints” that will be invaluable for autonomous chemistry.
Full article available here.