The Forensic Comparison of Soils by Automated Scanning Electron Microscopy

Abstract

ABSTRACTIn this preliminary study, the authors have developed a soil analysis method based on the automated particle search and classification capabilities of a scanning electron microscope equipped with an energy dispersive x-ray spectrometer (SEM-EDX). Innovative sample preparation maximized dispersion and density of soil particles while yielding highly representative subsamples. The automated SEM-EDX scheme, while analogous to traditional soil analysis by optical mineralogy, does not suffer from some of the serious disadvantages associated with optical methods. The analysis of soils by SEM-EDX is fast and reliable, requiring little intervention by the examiner once mineral categories have been defined, whereas analysis by optical mineralogy is extremely time-consuming and demands the skill of an examiner highly trained in the use of a pétrographie microscope. Automated analysis not only classifies soil particles much faster than a human analyst, but also allows an improved objective statistical measure of soil similarity or difference. The ability of the automated system to successfully discriminate different soils, yet to ‘recognize’ replicate samples, was successfully demonstrated. As a further advantage, automation permits archiving of the raw data for re-interpretation as new mineral categories are added to the classification scheme. The preliminary results indicate that automated analysis provides amples discrimination as good as, or better than, can be achieved by optical mineralogy.