Image Analysis Optimization for Quantifying Nanoparticle Dispersions in Polymer-based Nanocontposites Using Transmission Electron Microscopy (TEM)

摘要

When comparing large numbers of TEM micrographs of insoluble additives in polymer-based nanocomposite systems, the ability to determine or estimate the dispersion quality (i.e. uniformity of size and/or spatial distribution) is often difficult. The objective of this study was to develop a method to quantify dispersions observed in TEM micrographs that enables both a numerical "ranking" to be assigned to individual dispersions as well as tabulation a multitude of images acquired over time. Several methods were reviewed and applied to a set of TEM dispersion images of an insoluble additive in polystyrene. Projected area diameter, particle area, and Euclidean distance between particle centroids were chosen from all the particle size distribution and spatial distribution parameters present in the literature, but none successfully yielded a quantitative indicator of dispersion quality for the micrographs. In contrast, generating cumulative volume percent curves for each sample appeared to be a preferred method of quantifying and comparing dispersions in TEM micrographs. The volume diameter values obtained by this method can be used for "ranking" and tabulation of dispersion quality and account for both "good" additive dispersions (i.e. those with small domains of a narrow size range around 1 μm or less) and "bad" additive dispersions (i.e. those with non-uniform domains ranging in size by several microns or more). As a result, the numerical values generated by this method can be used to quantitatively determine correlations between the dispersion quality of nanoparticles in polymer-based nanocomposite materials and various macroscale physical and/or performance properties of such materials. This method's precision was statistically determined to decrease with increasing particle size and be heavily dependent on representative sampling.