Multimodal spectroscopy with chemometrics: Application to simulated forensic soil casework

Abstract

A multimodal analytical sequence in combination with chemometrics was applied to a blinded case simulation to showcase its suitability for use within the context of forensic casework involving arid, sandy soils low in clay and organic matter. Using microspectrophotometry, attenuated total reflectance Fourier-transform infrared spectroscopy and X-ray diffraction, in combination with principal component analysis and linear discriminant analysis, an unknown “recovered” sample was able to be associated with a single reference soil and discriminated from the other soils. The spectroscopic techniques employed produced varying results, however, these results offered complementary information that was able to be interpreted collectively to discriminate between the reference soils and identify the source of the “recovered” soil with greater confidence. The use of chemometric methods to interpret the data provided an objective, statistical measure of the similarities and/or variation between samples that could be presented as weighting for evidence in court. Hence, this approach was deemed suitable for incorporation into forensic casework to provide discrimination between similar sandy soil samples in an objective manner.