Advanced Research in Microspectrophotometry of Fibers: Analysis and Interpretation

Microspectrophotometry (MSP) is one of the most established instrumental methods used in trace evidence comparisons, and it is used by nearly every laboratory that conducts trace evidence analysis. Given this history, it may be expected that all significant avenues of research have been explored; however, through our own casework involving the analysis of thousands of fibers, our laboratory's internal research, and literature reviews, we have identified numerous, unaddressed topics that each have a critical impact on collection and interpretation of MSP data. These topics relate to spectral quality, spectral discrimination, and the efficacy of ancillary methods. The research proposed is designed to address these topics with the ultimate goal of formalizing the match criteria used to discriminate fibers. Part I - Spectral Quality. Prior to evaluating match criteria, it is critical that variables impacting spectral quality are controlled as closely as possible. Currently, many significant factors impacting data collection go unnoticed and/or uncontrolled. Some of the topics addressed include: the impact of edge contrast due to differences in refractive index of fiber and mounting medium; the impact of delustrants and cross section on reproducibility; treatment of heavily, lightly and heterogeneously dyed fibers; and sample alteration from illumination induced bleaching. Each of these factors is frequently encountered, has a significant impact on the resulting spectrum, and goes largely uncontrolled, making critical spectral comparisons impossible. The research here will systematically examine the effects of these factors and result in the establishment of procedures to actively address them in casework. Part II - Spectral Discrimination. Spectral comparisons conducted in casework are almost entirely visual. This approach is not systematic, does not take full advantage of the analytical data, and offers no formal way to evaluate the significance of subtle spectral differences when drawing conclusions. While MSP research has presented various statistical methods for spectral comparison, none have been practically applicable to casework on a regular basis. In contrast, the glass community has recently developed a successful and systematic approach to match criteria in casework. Using high quality spectra collected under the methods developed in Part I, carefully selected populations of fibers will be analyzed from our collection of over 2000 fabric reference samples, which have been colored with known dyes of known strength on known fibers. These fibers will be used to explore and evaluate match criteria based on existing comparison methods including visual overlay, 1st derivative, range plus standard deviation, and PCA. Part III - Efficacy. Most MSP systems are equipped at least some of the following accessories: UV (250-400 nm), NIR (700-900 nm), and Fluorescence. While scattered studies have addressed specialized applications, no formalized guidance exists to assist examiners in determining if any of the above provide increased discrimination, and if so, under what circumstances they are applicable in casework. The results of Parts I and II, our reference collections, and our practical experience in fiber examination will be used to direct a formal evaluation of these methods. In summary, the overall research plan is based on the unique combination of our staff's casework and research experience, existing equipment, and arguably the most extensive dyed fiber, fabric and colorant reference collections in the world. Results will be published in a "Practical Guide to Methods and Interpretation in MSP Analysis" with procedures and validation studies suitable for integration into ISO 17025 compliant procedures. On a larger scale, the results of this research will serve to advance the scientific rigor of MSP data collection and interpretation, specifically in light of the NSF study. ca/ncf