|
1 Camborne School of Mines, School of Geography, Archaeology and Earth Resources, University of Exeter, Redruth, Cornwall TR15 3SE, UK dpirrie{at}csm.ex.ac.uk
2 CSIRO Minerals, Queensland Centre for Advanced Technologies, Technology Court, Pullenvale, QLD 4069, Australia
3 Intellection Pty Ltd, Milton, Brisbane, Queensland, QLD 4064, Australia
QemSCAN is a scanning electron microscope (SEM) system, initially designed to support the mining industry by providing rapid automated quantitative mineral analyses. The system is based upon Carl Zeiss SEMs fitted with up to four light-element energy dispersive X-ray spectrometers. Representative subsamples are mounted into either resin or wax blocks and polished prior to analysis, or can be mounted onto carbon tape. During analysis, X-ray spectra are collected at a user-defined pixel spacing and are acquired very rapidly (c. 10 ms per pixel). The measured spectra are automatically compared against a database of known spectra and a mineral or phase name is assigned to each measurement point by the QemSCAN computer software programs. In this way the near-surface qualitative elemental composition of each particle is systematically mapped, assigned to a mineral name or chemical compound/species, and digital pixel maps of each particle are created. Depending upon a range of parameters, including the particle size and the user-defined pixel spacing (which can vary between 0.20 µm and 25 µm), approximately 1000 particles, each 1–10 µm in size, can be measured per hour using a 1 µm pixel spacing. In addition to providing a qualitative elemental analysis and mineralogical or phase assignment for each particle, data relating to particle size, shape and calculated specific density are also generated. In this study, the potential application of this automated SEM system in forensic geoscience was evaluated by the analysis of: (1) a series of soil samples, and (2) a series of dust samples from an industrial complex. In both case studies, the mine-ralogy/phase composition of each sample analysed was found to be distinctive. In addition, textural data for the soil samples and particle shape data for the dust samples show that they can be clearly distinguished. Automated SEM using QemSCAN has clear potential application in the analysis of soil or other trace evidence in forensic case work.
