The knowledge of product particle size distribution (PSD) in crystallization processes is of high interest for the pharmaceutical and fine chemical industries, as well as in research and development. Not only can the efficiency of crystallization/production processes and product quality be increased but also new equipment can be qualitatively characterized. A large variety of analytical methods for PSDs is available, most of which have underlying assumptions and corresponding errors affecting the measurement of the volume of individual particles. In this work we present a method for the determination of particle volumes in a bulk sample via micro-computed tomography and the application of artificial intelligence. The particle size of bulk samples of sucrose were measured with this method and compared to classical indirect measurement methods. Advantages of the workflow are presented.
Keywords: Mask–RCNN; Wadell sphericity; determination of particle size distributions; micro-computed tomography; three dimensional particle analysis in bulk.