Date of Award
Campus Access Dissertation
Chemistry and Biochemistry
Stephen L. Morgan
The objective of this dissertation is to provide the scientific foundation for spectroscopic detection of biological fluids at crime scenes. Identifying these fluids using spectroscopy may replace existing presumptive chemical-based tests that have disadvantages (i.e., toxicity and sensitivity to light). The feasibility of remote spectroscopic detection to visualize blood on forensically relevant textiles was evaluated using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). Reflectance spectra of bloodstains contain distinctive amide bands due to the proteins in blood. Amide I, II, and III bands (1650, 1540, and 1200-1350 cm-1 respectively) were used to differentiate between neat and blood-doped substrates.
The forensic applicability of a direct spectroscopic method for estimating the age of bloodstains is also described in this research. DRIFTS was utilized in dating dried blood on textiles made possible by the changes in the relative amounts of secondary protein structures as blood oxidizes and degrades with time. These changes were modeled as a function of time with principal component regression (PCR) while interval PCR (iPCR) was used to locate the optimal spectral regions associated with the changes due to blood aging.
The sampling and collection of spectra was evaluated in order to decrease the spectral variation and enhance the discriminating power. High discrimination power is important so that time and resources of the forensic investigators are not wasted on collecting and analyzing false positive samples. Doped samples were dip coated to create a more uniform application. Using DRIFTS again, spectra of neat and doped textiles were acquired using a controlled orientation. The improved reproducibility allowed principal component (PCA) and linear discriminant analysis (LDA) to be more effective in discrimination. Furthermore, PCR was used to show that detection limits for blood were greatly decreased.
Textile fibers are significant pieces of trace evidence recovered from crime scenes. Both UV/visible microspectrophotometry and attenuated total reflectance (ATR) spectroscopy were used to analyze the fibers. PCA and LDA were employed to test the ability of providing additional discrimination between groups of fibers by calculating the first derivative. The classification accuracies along with the false positive/negative rates were compared to examine the effect of applying the first derivative to the spectra.
McCutcheon, J. N.(2010). Forensic Discrimination, Age Estimation, and Spectral Optimization For Trace Detection of Blood On Textile Substrates Using Infrared Spectroscopy and Chemometrics. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/374