Commercially available domestic and imported red-dyed cotton fabrics were examined using conventional techniques (see Appendix B) to determine fiber color and dye classification. Color was analyzed by Microspectrophotometry (MSP); the discriminating power of this technique was evaluated. Dye class was established by standard extraction protocols (see Appendix C). Reproducibility of dye classification by extraction for the samples was evaluated. When dye-classification and color determination by MSP were insufficient to provide discrimination between fibers from different sources, novel techniques were employed. Liquid chromatography-mass spectrometry analysis was conducted to discriminate fibers with matching dye class and MSP spectra. Ultraviolet fluorescence spectroscopy was employed and evaluated as an alternative and as a supplement to LC-MS to further reduce the chance of false positive associations.
cteristic used for comparing cotton fibers. Microspectrophotometry (MSP) generates spectral profiles to international standards (CIE), having a dominant wavelength (? max) that defines the principal color of a dyed fiber. Absorption spectra generated by MSP analysis can categorize fibers into spectrally matching pairs. Chemically, different dyes of the same color class can yield matching spectral profiles, generating false matches in approximately 5% of the casework (Hartshorne & Laing, 1988; Fong & Inami, 1986).
Microspectrophometry is a method in which a dyed fiber on the stage of a light microscope is illuminated with incident white light. Individual wavelengths are absorbed corresponding to the specific molecular structure of the fiber/dye while the remaining wavelengths are transmitted to a detector to allow the generation of a signature absorption spectrum.
Liquid chromatography-mass spectrometry (LC-MS) is used to separate and identify dyes extracted from colored textile fibers. 'The combination of liquid chromatography and mass spectrometry (LC-MS) is a highly selective and sensitive analytical method that has been shown to be useful for the characterization of dyes according to their molecular structures' (Huang, Yinon, & Sigman, 2004, p. ). As dye components elute through the HPLC column they undergo electrospray ionization (ESI). 'Electrospray ionization (ESI) is arguably the most universal technique for the ionization of nonvolatile and thermally-labile molecules for mass spectrometry analysis' (Pamanik, Ganguly, & Gross, 2002, p. ). This soft ionization technique tends to generate molecular ions and create multiple charging on the fragments, which allows the determination of mass for large molecules to be accomplished without a specialized detector or a mass spectrometer with an extended m/z range. 'Modern LC-MS methodologies are available that utilize very low flow rates (mL/min even nL/min rates) and thus require extremely low sample volumes and are highly congruent with the needs of forensic analysts' (Tuinman, Lewis, & Lewis, 2003, p. ). According to Huang et al (2004), 'Extraction of textile dyes almost invariably leads to the recovery of additional, often UV-absorbing, components such as fluorescent brighteners. These fiber components would also be important to include in a database designed to meet the needs of the forensic community' (p. ).
Fluorescence Spectroscopy utilizes ultraviolet light to irradiate a sample that contains fluorescent material. Fluorescent materials are organic molecules containing conjugated double bonds are capable of absorbing UV. Fluorescence is the phenomenon in which the fluorescent molecule emits a longer wavelength upon decay as was absorbed during excitation. The distribution of the intensities of the absorbed wavelengths is known as the fluorescence excitation spectrum, whereas the intensity distribution of emitted wavelengths is known as the fluorescence emission spectrum. Excitation characteristics of dyes and brighteners can be observed in addition to effects caused by fabric conditioners. The graph of the fluorescence intensity as a function of the excitation and emissions wavelengths (EEM) yields information about the sample composition. These treatments in combination with local environmental contaminates transferred by contact may make each fiber unique to its source. However, the EEM spectra generated are a result of the combined fluorescent effects of dyes, brighteners, fiber processing, and the raw cotton itself. While a questioned-known match can be established using fibers that are not altered between deposition and collection, the effect of laundering a garment might alter the spectral profile enough to generate a false negative result if fluorescent materials were transferred to the surface of the fiber from the detergents.
A SEE 2100 microspectrophotometer with a xenon lamp was used to produce the absorption spectra for each analysis. A background scan and a dark scan can be performed to correct for the spectral power distribution from the light source and to eliminate any environmental interferences.
An Agilent 1100 MSD quadrupole mass spectrometer equipped with an electrospray ionization (ESI) source and interfaced to an Agilent 1100 HPLC was used for this study. This instrument can operate in either positive or negative ion mode for the detection of dyes that form either positive of negative ions. Fragmentor voltage can be adjusted to alter the extent of molecular decomposition.
A customized bench-top Photon Technology International Ultraviolet Spectrofluorometer was used to measure excitation wavelengths of the extracted dyes/brighteners. A continuous wave xenon lamp was used as the source. Solvent fluorescence can be subtracted using a background scan.
Water was prepared from Barnstead E-pure (Barnstead/Thermolyne, Iowa), with a resistance of 18.2 megohm-cm. Pyridine (Aldrich Chem. Co. Milwaukee, Wisconsin), methanol (HPLC grade, AlliedSignal Inc., Burdick & Jackson, Muskegon, MI), dimethylformamide (Aldrich Chem. Co. Milwaukee, Wisconsin), glacial acetic acid (Aldrich Chem. Co. Milwaukee, Wisconsin), dithionite (Aldrich Chem. Co. Milwaukee, Wisconsin), polyvinylpyrolidone (Aldrich Chem. Co. Milwaukee, Wisconsin), and 12% sodium hydroxide (Aldrich Chem. Co. Milwaukee, Wisconsin) were used to prepare the extractant solutions.
Five solutions were prepared for dye-class determination by solvent extraction: Pyridine/Water (4:3) ; Dimethylformamide(DMF)/Water (3:1); Glacial acetic acid used as received; Dithionite/Polyvinylpyrolidone (PVP) (1:1); and 12% Sodium hydroxide. Threads 1 cm in length were chopped into smaller pieces to increase surface area and extracted with 0.5ml solvent in sealed glass tubes fashioned from pasture pipets. Heating varied according to the conditions specified in the protocols.
Two solutions were prepared for dye extraction for LCMS: Pyridine/Water (4:3); Methanol/Water (1:1). A 5mm piece of thread was extracted with a 20 'L aliquot of solvent in a sealed glass capillary under heating at 150'C. HPLC-MS analysis was made on 5 'L injections.
Extractions prepared with 3:1 DMF extractant and extractions prepared with a 1:1 solution of ethanol/water extractant were tested in order to determine the solvent that would best extract the dye without interfering with the fluorescence profile. Ultimately, DMF was exhibited undesirable background fluorescence. Ethanol provided the optimal dye extraction with low background fluorescence. A 5mm piece of thread was extracted with a 20 'L aliquot of solvent in a sealed glass capillary under heating at 150'C.
Textile fibers were obtained from commercially available fabrics. Threads 1 cm in length were chopped into smaller pieces to increase surface area and extracted with 0.5 mL solvent in sealed glass tubes fashioned from pasture pipets. Heating varied according to the conditions specified in the protocols. Parallel extractions were carried out for each sample in each solvent. An aliquot of each sample was extracted first with glacial acetic acid at 100'C for 20 min. If a good extraction was not obtained a new thread from the same sample was exposed to aqueous pyridine at 100'C for 20 min. If a good extraction was not obtained another aliquot was extracted with a dithionite and PVP mixture at 100'C for 20 min. Extracts were then spotted on a C18 TLC plate and illuminated with UV light of ? = 245nm. If no dark spot appeared, indicating no absorbent analyte present, the fiber was examined for a color change. If a dark spot appeared, indicating an absorbent analyte present, a new fiber was extracted with 12% NaOH at 100'C for 20 min. The fiber was then examined for color change.
A fiber was removed from the thread and placed on a glass slide. A glass rod was used to flatten the fiber yielding a more uniform thickness. No cover slip was atop the fiber. Spectra were obtained from 5 sites along the length of each of 5 fibers from a textile thread, generating a total of 25 spectra per sample. The spectra for each sample were averaged to yield 10 spectra, one for each textile. These averaged-spectra were normalized by area across a 400-700nm-wavelength range for comparison. A reference scan and a dark scan were conducted to correct for background and instrumental interferences. The parameters were set for an average of 100 scans per data point. The graph of the absorption spectrum was limited to the visible range 400 to 700 nm.
Separation was carried out on a ZORBAX Eclipse XDB-X18 (2.1 x 150 mm) HPLC column at a mobile phase flow rate of 0.20 mL/min. A programmed solvent gradient (methanol/water) was used to achieve better separation. A 1:1 methanol/water mixture was held constant for the first 5 min of analysis, followed by a steady increase in the methanol composition to 95% at 25 min. The methanol composition was held constant at 95% until the analysis ended at 40 min. Long elution times for each run are necessary to elute all of the co-extracted components and keep the column clean to ensure reproducible performance. Mass spectrometer parameters were optimized for maximum sensitivity. The drying gas for the ESI was held at 12.0 L/min and the spray chamber temperature was set at 350'C for all analytes, unless otherwise specified.
The sample extract was transferred to a quartz cuvette for analysis. Electron excitation behavior was graphed as a function of monochromatic wavelength. The emission matrix was produced from a scan across a wavelength range of 325-800nm, while the excitation matrix was produced from a scan across 285-335nm. Slit settings were 6.00nm for excitation and 6.00nm for emission respectively.
View PDF Rachel Russo is a senior in the Burnett Honors College enrolled in the Forensic Science program at the University of Central Florida. Her research concentrates in applied analytical chemistry for the investigation of fiber dyes. This research was conducted over the course of two years at the National Center of Forensic Science and the University of Central Florida Chemistry Department as a project for the Burnett Honors College Honors in the Major Program, under the guidance of Dr. Barry Fookes with assistance from Dr. Michael Sigman and Dr. Andres Campiglia. Upon graduation Rachel plans to attend medical school to pursue a career in forensic radiology.
© 2007 UNIVERSITY OF CENTRAL FLORIDA UNDERGRADUATE RESEARCH JOURNAL -COMPUTER CENTER II ROOM 209- ORLANDO, FL 32816-0086 - (407) 823-3125
Office of Undergraduate Research
Contact the webmaster with website questions.