Forensic anthropologists normally examine bone from a variety of medicolegal contexts. The skeletal remains may in some cases be highly fragmented or taphonomically modified, making it difficult to sort bone from non-bone material. In these cases, the forensic anthropologist may rely on microscopic or destructive chemical analyses to sort the material. However, these techniques are costly and time-intensive, prompting the use of nondestructive analytical methods in distinguishing bone and teeth from non-bone materials in a limited number of cases. The proposed analytical techniques are limited in that they rely on an examination of the major elements in the material, and do not sort out all materials with a similar chemical composition to bone/teeth. To date, no methods have been proposed for the use of handheld Xray fluorescence (HHXRF) spectrometry in discriminating human and nonhuman bone/teeth from non-bone materials. The purpose of this research was to develop a method for the use of HHXRF spectrometry in forensic anthropology specifically related to distinguishing human and nonhuman bone and teeth from non-bone materials of a similar chemical composition using multivariate statistical analyses: principal components analysis (PCA), linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and hierarchical cluster analysis (HCA). This was accomplished in two phases. Phase 1 consisted of a Reliability Test and involved sampling a single human long bone in thirty locations. Multiple spectra were collected at each location to examine the reliability of the instrument in detecting the elements both within a single site and between multiple sites. The results of the Reliability Test indicated that HHXRF consistently detected the major and minor elements found on the surface of a human bone. iv These results were used for Phase 2, designated the Accuracy Test, which involved analyzing a set of materials compiled from the literature to test the accuracy of the technique in discriminating bone (human and nonhuman) and non-bone samples (other biological and nonbiological). The results of the Accuracy Test indicate that osseous and dental tissue can be distinguished from non-bone material of similar chemical composition with a high degree of accuracy (94%) when data is collected from several locations on a sample and analyzed separately during multivariate statistical analyses. Overall, it was not possible to discriminate rock apatite and synthetic hydroxyapatite (synthetic bone) from bone. However, this technique successfully discriminated other non-bone materials that are chemically similar to bone, such as ivory and octocoral, which previous methods focusing on only a comparison of Ca/P ratios were unable to distinguish from bone.

Master of Science / Geology / Matthew W. Totten / Organic-rich shales were deposited over a large part of what is now North America during the Late Devonian. North America in the late Devonian was located in the tropics (Woodrow et al., 1973), possibly in low southerly latitudes (Heckel and Witzke, 1979; Witzke and Heckel, 1988; Streel et al., 1990). This environment creates an organic-rich environment that resulted in thick, black shales. The Devonian-Mississippian Chattanooga (Woodford) shale is known to be an important petroleum source rock in many intracratonic basins of the Midcontinent (Lambert, 1993). Geochemical analysis of the Chattanooga shale, using various techniques, provides additional information on oil-source rock potential. Handheld XRF analysis was conducted on well cuttings samples, Loss on Ignition (LOI) was performed on a subset of those samples and mapping of the organic matter results of the two methods was completed. Handheld XRF still has the prospect for providing quick analysis to infer organic matter content to be used as a determination of the quality of source rock. Although slightly reduced correlation has been found within this study compared to Willey (2015), the method has still proven viable for fracking targets to be determined on site and in a more efficient manner. Loss on Ignition results have correlated with TOC data better then XRF results, making this method the better option for evaluating source rock potential. Mapping of these results provide the first known source rock potential map across Kansas and can be used by the industry for future exploration.