Shape identification and classification of aerosol particles using SEM image

Abstract

For the analysis of aerosol, the bulk analysis method of collected particles on a filter is widely used. In the bulk analysis, the information with high accuracy as the averaged value of collected particles can be obtained, but the information of individual particle is disappeared. Some surface analysis techniques such as SIMS and SEM-EDX have also been employed for individual particle analysis. Since the individual analysis for all the collected particles is not realistic in terms of analytical time, statistically reliable data cannot be obtained. Therefore, it is difficult to compare the data of individual particle analysis with bulk analysis data and discuss the abundance ratio of particles. In this study, aerosol was analyzed with a high resolution by means of FIB-TOF-SIMS developed by our group. From the results of the individual particle analysis, it was found that there is a good correlation between the aerosol type and shape of particles. But, these need much time and labor. Therefore, a method for classification particles from shape data was devised. Based on a shape to be provided from an SEM image, they are classified by means of cluster analysis.