High‐throughput and High‐speed Absorbance Measurements in Microfluidic Droplets using Hyperspectral Imaging

The recent progress of machine learning and microfluidics in the chemical and biological sciences has motivated the development of new online techniques to interrogate the (bio)chemical contents within moving droplets. To accelerate the optical characterization of new materials and chemical reactions, we combine a line‐scan hyperspectral imaging system with a droplet‐based microfluidic reactor. We demonstrate the performance of this platform on a model chemistry – silver nanoparticle synthesis. The platform can image the spectral signature of ∼400 individual droplets in only 15 s, with droplet flow speeds exceeding 4 cm/s in the reaction tube. After correction of the keystone and smile effects on the hyperspectral images, the absorbance spectra are extracted from the droplets with an accuracy comparable to industrial spectrophotometers. The time evolution of the UV/Vis absorbance spectra during the reactive synthesis can be tracked either by scanning all the droplets present in the reaction tube or by following a subset of the droplet ensemble at frame rates up to 92 fps. This high‐throughput and high‐speed platform is particularly interesting for screening large parameter spaces and imaging fast reactions with a high resolution, for eventual coupling with advanced machine learning techniques to infer kinetic models and obtain detailed mechanistic insights.