Automated determination of aggregate primary particle size distribution by tem image analysis: Application to soot

Abstract

In many applications, nanoparticles appear to be in an aggregated form. Thus, a complete description of their morphology involves an analysis of their size at different scales, from the aggregate size to the primary particle size. In this study, we present an automated method for the determination of the primary particle size distribution. It is based on an image-processing algorithm operating Euclidian distance mapping leading to a function that contains information about aggregates morphology. This algorithm is first applied to virtual aggregates with point contact between spheres, generated by diffusion limited cluster aggregation code. It shows that obtained functions can be used to retrieve the parameters of the primary particle diameter size distribution. However, it also demonstrates that overlapping or necking effects between primary spheres can have an impact on the shape of the function. The analysis is also performed on real soot images generated by ethylene diffusion flame and civil aviation engine that are different from a physicochemical point of view. This leads to the definition of a calibrated morphological function that can be used for the determination of the parameters of the primary diameter distributions of real soot particles. Finally, the same method is applied to TEM images of soot generated by a commercial soot generator miniCAST. Three operating conditions were investigated showing that generators can be used to produce soot particles with morphologies similar to those produced by jet engines.Copyright 2014 American Association for Aerosol Research © 2014 Copyright © American Association for Aerosol Research.