Automated forensic soil mineral analysis; testing the potential of lithotyping

Abstract

AbstractIn the investigation of serious crimes, soil can be, in some cases, a very valuable class of trace evidence. The complexity of soil is part of the reason why it is useful as trace evidence but is also an inherent problem, as there are many different parameters in a soil sample that could potentially be characterized. The inorganic components of soils are dominated by minerals, along with anthropogenic particulate grains; thus, the analysis of soil mineralogy as the main technique for inorganic forensic soil characterization is recommended. Typical methods that allow the bulk mineralogy to be determined, such as X-ray diffraction (XRD), do not allow the texture of the particles to be characterized. However, automated scanning electron microscopy (SEM) provides both modal mineralogy and also allows particle textures to be characterized. A recent advance in this technique has been the ability to report the modal mineralogy of a sample as ‘lithotypes’, which are defined on the basis of a combination of mineralogy and other parameters, such as grain size and mineral associations. Defined lithotype groups may include monominerallic grains but also, importantly, allow the automated quantification of rock types and other anthropogenic materials. Based on a simulated forensic scenario, the use of lithotyping is evaluated as an aid in the analysis of soil samples. This technique provides additional discrimination when comparing different soil samples.