Computer Vision for Kinetic Analysis of Lab- and Process-Scale Mixing Phenomena

A software platform
for the computer vision-enabled analysis of
mixing phenomena of relevance to process scale-up is described. By
bringing new and known time-resolved mixing metrics under one platform,
hitherto unavailable comparisons of pixel-derived mixing metrics are
exemplified across non-chemical and chemical processes. The analytical
methods described are applicable using any camera and across an appreciable
range of reactor scales, from development through to process scale-up.
A case study in nucleophilic aromatic substitution run on a 5 L scale
in a stirred tank reactor shows how camera and offline concentration
analyses can be correlated. In some cases, it can be shown that camera
data hold the power to predict reaction progress.