Elemental fingerprinting of soils using ICP-MS and multivariate statistics: A study for and by forensic chemistry majors

Abstract

Students in an instrumental analysis course with a forensic emphasis were presented with a mock scenario in which soil was collected from a murder suspect's car mat, from the crime scene, from adjacent areas, and from more distant locations. Students were then asked to conduct a comparative analysis using the soil's elemental distribution fingerprints. The soil was collected from Lafayette County, Mississippi, USA and categorized as sandy loam. Eight student groups determined twenty-two elements (Li, Be, Mg, Al, K, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Rb, Sr, Cs, Ba, Pb, U) in seven samples of soil and one sample of sediment by microwave-assisted acid digestion and inductively coupled plasma-mass spectrometry (ICP-MS). Data were combined and evaluated using multivariate statistical analyses. All eight student groups correctly classified their unknown among the different locations. Students learn, however, that whereas their results suggest that the elemental fingerprinting approach can be used to distinguish soils from different land-use areas and geographic locations, applying the methodology in forensic investigations is more complicated and has potential pitfalls. Overall, the inquiry-based pedagogy enthused the students and provided learning opportunities in analytical chemistry, including sample preparation, ICP-MS, figures-of-merit, and multivariate statistics.