This article reports on the development of a method for the quantitative elemental analysis of surface soil samples using laser-induced breakdown spectroscopy (LIBS) and its application to the analysis of bulk soil samples for discrimination between specimens.
The use of a 266 nm laser for LIBS analysis is reported here for the first time in forensic soil analysis. Optimization of the LIBS method is discussed, and the results compared favorably to a laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) method previously developed. Precision for both methods was <10 percent for most elements. LIBS limits of detection were <33 ppm and bias <40 percent for most elements. In a proof of principle study, the LIBS method successfully discriminated samples from two different sites in Dade County, FL. Analysis of variance, Tukey’s post hoc test and Student’s t test resulted in 100 percent discrimination, with no type I or type II errors. Principal components analysis (PCA) resulted in clear groupings of the two sites. A correct classification rate of 99.4 percent was obtained with linear discriminant analysis using leave-one-out validation. Similar results were obtained when the same samples were analyzed by LA-ICP-MS, showing that LIBS can provide similar information to LA-ICP-MS. In a forensic sampling/spatial heterogeneity study, the variation between sites, between sub-plots, between samples, and within samples was examined on three similar Dade sites. The closer the sampling locations, the closer the grouping on a PCA plot and the higher the misclassification rate. These results underscore the importance of careful sampling for geographic site characterization. (publisher abstract modified)