Results and Discussion

The results confirm unique identification of all 10 fibers using these additional investigative tests, a task not possible by conventional analysis alone. Analysis by multiple techniques greatly enhances the probative value of trace fibers in criminal investigations by providing fiber discrimination at a higher degree of certainty. The application of liquid chromatography-mass spectrometry (LC-MS) and fluorescence spectroscopy to the analysis of cotton fibers is shown to greatly increase their evidentiary value by providing highly specific chemical and structural information about the dyes and brighteners.

Each of the 10 samples was readily distinguishable in the bulk samples. However, when comparing individual fibers visually or microscopically no distinction could be made. Dye extraction alone provided inconsistent results. Often the fibers would react with more than one extraction solution. Perhaps there were multiple dyes used on a single fiber. The evaluation criteria in the protocol, to establish whether a 'good extraction' was obtained, was subjective and lead to different dye classifications for the same sample depending on the analyst and the trial. Dye extraction results were inconsistent and not adequately reproducible. After twelve independent analyses the dyes were classified as listed in Appendix D. Dye extraction alone was able to differentiate only 2 out of 10 samples.

The samples could be divided into two groups based on MSP absorbance data: those exhibiting a single peak, and those exhibiting a double peak. Within the second group, sub-classes could be determined according to peak ratios. In 6 out of 10 cases MSP was unable to differentiate between two or more fibers that were visually distinguishable in the bulk sample. In these cases, the dyes were extracted and their chemical natures were examined by LC-MS. The resulting mass spectra were set to the same scale for comparison. The chemical components of each dye were analyzed.

The 6 samples examined by LC-MS could be initially divided into 3 groups based on the characteristics of their MSP absorption spectra: group A consisted of fibers from samples 6 and 8, group B consisted of fibers from samples 3 and 4, and group C consisted of fibers from samples 1 and 2. Peak maxima were determined electronically by the SEE software accompanying the microspectrophotometer. The MSP spectra for group A fibers exhibited two maxima, one at 523nm and one at 554nm; the peak at 523nm was of greater intensity than the peak at 554nm. The MSP spectra for group B fibers exhibited two maxima, one at 526nm and one at 552nm; the peak at 552nm was of greater intensity than the peak at 526nm. The MSP spectra for group C fibers exhibited two maxima, one at 523nfsm and one at 550nm; the peaks were at near equal intensity with the peak at 523nm being only slightly greater than the peak at 550nm (see Appendix E).

Group A fibers were distinguishable by dye extraction behavior (see Appendix D). Sample 6 was classified as a reactive dye while sample 8 was classified as an azoic dye. LC-MS analysis provided readily distinguishable spectra. The extract from sample 8 was thermally unstable, decomposing with methanol extraction at 150' C and exhibiting a color change from red to yellow. The chromatogram indicated the dye(s) extracted from sample 8 consisted of 2 components, one with a pair of peaks at m/z 383 and 385 and the other with peaks at m/z 171 and 173. The mass spectrum also indicated the presence of chlorine. Sample 6 had none of these characteristics (see Appendix F).

Group B fibers were indistinguishable by dye extraction. However, they exhibited clearly discerning behavior upon MS fragmentation. While the retention times were similar (3.29min for sample 3 and 3.27min for sample 4), the extract of sample 4 favored negative ion current while the extract of sample 3 favored positive ion current. The MS spectrum of sample 3 indicated the probable presence of multiple chlorines, while no such indication was found in the spectrum of sample 4 (see Appendix F).

Group C fibers were also not distinguishable by dye extraction. The LC-MS spectra for samples 1 and 2 indicated that each contained two sulphonated sodium groups as both contained an M-23 and an M-46 peak. The spectrum for sample 1, however, indicated the presence of a halogen. The isotopes were too high in m/z ratio to be chlorine. Bromine was suspected (see Appendix F).

In order to further differentiate between Group C fibers samples 1 and 2 were subjected to UV Fluorescence Spectroscopic analysis (see Appendix G). Spectra unique to fibers from samples 1 and 2 when examined in situ were readily distinguishable and the spectra were reproducible. Additionally, extractions in ethanol produced identifiably unique, reproducible spectral profiles for each of the samples. While fluorescence spectroscopy is sufficient to provide reproducible results that easily aid a questioned-known match, more research needs to be done. Since the EEM spectra result from compounded fluorescent effects it is necessary to identify possible sources of variation. It was found that raw cotton fluoresces; however, it is not known if fiber maturity or cotton species/growing conditions alter the spectral profile. Since fibers are homogenized during the creation of a textile, there will be inherent variation within a single garment. Furthermore, fiber processing such as permanent press or mercerization could likely affect the EEM characteristics as both involve chemically and physically changing the fiber during processing. Changes in the chemical structure of the fiber and dye may alter the conjugated region of the fluorescent molecules. All the fibers in a garment have likely undergone the same processing so the effects should be universal; however, the same elements that complicate dye absorption affect the uniformity of processing. Uneven distribution of fibers or a concentration gradient of chemicals in a vat could lead to some fibers absorbing more chemicals than others, thus the chemical structures contributing to fluorescence spectra could be unequally affected. The regions of high dye concentration could result in excited dimmers which may have a different spectrum then for the same dye at lower concentrations. Furthermore, environmental factors that may have come into contact with the garment between its deposition of a fiber at the scene and its collection as evidence should be taken into account. Likewise, the effect of laundering a garment in varying detergents should be researched to determine the resultant changes in spectra from the transfer of fluorescent materials from the detergents to the fiber surface. It is not uncommon for suspect garments to be washed between the commission of a crime and their collection as evidence; variation as a result should not be misconstrued into a false negative.

The results of the preliminary data confirm unique identification of all the fibers using these enhanced investigative tests, a task not possible by conventional analysis alone. Multi-parameter analysis greatly enhances the probative value of trace fibers in criminal investigations by providing fiber discrimination at a higher degree of certainty. This study demonstrates the benefit of applying new techniques in the forensic investigation of fibers to reduce the chance of an incidental match. Sixty percent discrimination was achieved by employing current protocols; discrimination was improved to one-hundred percent by applying the methods outlined in this paper. Similarly, the application of liquid chromatography-mass spectrometry (LC-MS) and fluorescence spectroscopy to the analysis of cotton fibers is shown to greatly increase their evidentiary value by providing highly specific chemical and structural information about the dyes and brighteners.

There are many applications for LC-MS and fluorescence spectroscopy to detect trace chemicals on the surface of fibers. Future research into the detection of detergents and fabric conditioners would provide greater individualization. Two originally chemically identical textiles would possess matching dyed fibers directly after manufacture; however, the textiles (and fibers) would deviate chemically after the first laundering. Chemicals exclusive to specific brands of detergents or/and fabric softeners would be present in concentrations (and combinations) unique to each individual and should be detectable on the fibers. I have reason to believe, based on the research I have presented in my thesis, LC-MS and fluorescence spectroscopy are capable of such detection.

Some articles of interest on the effects of laundering on fiber investigations are listed below:

Bresee RR, & Annis PA. (1991). Fiber transfer and the influence of fabric softener. Journal of Forensic Science, 36(6), 1699-1713.

Palmer R. (1998). The retention and recovery of transferred fibers following the washing of recipient clothing. Journal of Forensic Science, 43(3), 502-504.

Lloyd JBF. (1977 0. Forensic significance of fluorescent brighteners: their qualitative TLC characteristics in small quantities of fiber and detergents. JFSS, 17, 145-152.

Bibliography and References >>