Performance quantification of liquid-liquid dispersion technology using an image analysis routine

The creation and control of interfacial area in liquid-liquid dispersions is key as it significantly influences intermediate and end product quality. This paper presents an image analysis routine based on the Hough transform to monitor droplet size distributions. The droplet size range was divided in three subranges allowing a tailored threshold for different circle sizes. The analysis routine was able to process 120 images per minute, resulting in the detection of 8000 circles per minute. Different weights of spherical glass beads were used as a validation. The detection performed well up till a glass bead concentration of 0.75 vol% and the relative standard deviations over three replicates (<5%) confirmed the reproducibility of the results. Manual verification showed that 65% of the undetected beads had a diameter below the detection limit. The Sauter mean diameter was obtained with 96% accuracy. Subsequently, the analysis routine was successfully applied to analyze images taken from samples from liquid-liquid dispersion experiments up to a volume 10 vol% dispersed phase. Samples were instantly stabilized and diluted by using a surfactant solution. Samples from a 50 vol% dispersion were successfully analyzed by implementing a protocol with a higher dilution factor. Facilitated by the improved analysis routine, a novel energy efficiency benchmark based on the Gibbs free energy was introduced. The benchmark aims to facilitate comparison over liquid-liquid systems with different interfacial tension and different power densities.