Validation of Forensic Characterization and Chemical Identification of Dyes Extracted from Millimeter-length Fibers

• NCJ Number: 248579
• Author(s): Stephen L. Morgan Ph.D.
• Date Published: September 2014
• Length: 142 pages
• Grant Number(s): 2010-DN-BX-K245
• Agencies: NIJ-Sponsored
• Publication Type: Research Paper; Report (Grant Sponsored)

Annotation

This research report describes the development and validation of methods for the forensic chemical identification and characterization of dyes extracted from trace evidence fibers.

Abstract

The objective of the research described in this report was to develop and validate methods for the forensic chemical identification and characterization of dyes extracted from trace evidence fibers. If the chemical composition of dye formulations on trace fibers can be reliably profiled by liquid chromatography (LC), match exclusions can be made with higher reliability and the results will have a solid scientific basis and increased practical significance. Separation and detection of individual dye components provides a qualitative and semi-quantitative fiber dye "fingerprint" that characterizes the number and relative amounts of dyes present. Chemical identity of extracted dyes can be inferred from LC retention time matching of dye peaks, comparison of UV/visible spectra of the separated dye components, and from molecular weight determination and structural analysis by mass spectrometry. A significant challenge faced by the researchers was that such analyses are destructive to the fiber evidence. Validation practices for calibration based on forensic and international standards were employed in this work and issues addressed with respect to estimation of limits of detection (LOD) included: 1) rational choices of measurement uncertainty are important for valid estimates of LOD; 2) reporting a range of multiple types of decision limits provides a stronger statistical basis for specification and control of false positive and false negative detection probabilities; and 3) use of Mandel sensitivity, which is independent of the scale in which measurements are expressed, as a useful tool for comparisons of variability between different analytical methods.