533 CHARACTERIZATION OF BLEACHED HAIR 9 reveals a steep decrease in normalized tryptophan with proportional increases in whole fiber cysteic acid. In contrast to whole fiber bleaching kinetics, oxidizing cystine in the cortical volume, which is encapsulated within the cuticle, was diffusion controlled, where the cuticular barrier initially mediated cortical absorption of alkaline peroxide. One possible explanation for the early plateauing of cortical FTIR-ATR data (first 60 min, Figure 8) is that the bleaching solution formed a viscous semisolid toward the end of each 1 h bleaching step, thereby hindering the rate of peroxide/persulfate cortical permeation. Consequently, the cuticle versus cortical EDF kinetical comparison for the FTIR-ATR studies (Figure 8 overlay), as judged by augmentation in the intensity of the normalized 1040 cm−1 cysteic acid band, is quite unambiguous. Initially, the cuticle received the brunt of the bleaching impact, whereas oxidative changes to the cortex were limited by tortuous permeation gateways introduced by cuticle and cortical diffusion networks. The disparity between the cuticle and cortical cysteic acid response is reconcilable when looking at each compartment of the hair fiber. While the morphology and molecular structure of the cuticle make it a formidable barrier for larger molecules, smaller molecules such as water, H 2 O 2 , and persulfates readily permeate the fiber structure, where diffusion control regulates further cortical permeation. Furthermore, excessive bleaching subsequently induces interfacial chemical attacks, leading to the formation of pores, cracks, and zones of erosion in the cuticle and cortex, thereby disturbing the natural diffusion processes and increasing the overall hydrophilicity of the fiber (23). As a result, trends in Figure 8 provide a “broad brush” for predicting subsequent physical properties that would be altered by cystine scission and the associated formation of cysteic acid. RAMAN SPECTROSCOPY OF CROSS-SECTIONED HAIR As the same cross-sectioning techniques for FTIR imaging and FTIR-ATR spectroscopic examinations were applied to the Raman studies, the high aspect ratio of the 3-µm thick cross-sections assured that the Raman spectra were predominantly composed of cortical and medullar shift frequencies. In the Raman experiments, instead of using European dark brown tresses, European natural white hair was employed because European dark brown hair contains melanin granules that induce excessive fluorescence, therein making it difficult to resolve Raman shifts for virgin or lightly-bleached dark brown fibers. Prior work in confocal Raman spectroscopy and multivariate curve resolution analyses provides insights toward assigning the spatial assignment of functional groups to either the cuticle or cortex (17,18). These spectral markers allow for some specificity of band assignments in the work reported by this laboratory. A table of published Raman active band assignments is presented in Table I. Works by both Kuzuhara, as well as Pudney et al., infer exclusivity of certain vibrational modes as a function of their placement in either the cuticle or cortex (17,18). Examples of such domains in the Pudney work include bands in the 650 cm−1 (-C—S-, gauche) and 880 cm−1 spectral regions of the cuticle, and conversely, the 742 cm−1 CH 2 in-phase and 1002–1003 cm−1 phenylalanine bands that prominently appear in the cortex. Of course, many vibrational bands discussed in this study are shared between physical regions, such as the 1451 cm−1 C—H bending and 1650–1675 cm−1 amide I bands. While Kuzuhara’s study agrees with Pudney on the specificity of some bands, including 1342 cm−1 (-CH 2 bend), he generally found the presence of most signature Raman bands, such as phenylalanine (1003 cm−1), in both domains of

534 JOURNAL OF COSMETIC SCIENCE the Raman hair map. In fact, the Kuzuhara results align with amino acid analysis work published by Robbins, which advocates that phenylalanine is present in both the cuticle and cortex (1). Although the bulk Raman methodology used in our studies does not offer the compartmental selectivity asserted in the confocal studies of Pudney, the work is in line with Kuzuhara. For example, the 1003 cm−1 phenylalanine band was present in the cuticle as well as the cortex, which is assumed to be the bulk of the cross-sectioned fiber. With some exceptions, both cortex spectra from cross-sections (assuming cuticle contributions do not exceed 7–8% of the total signal) and cuticles largely reflect their spatial origination with superior signal-to-noise ratios over their confocal Raman microscopy counterparts. With the reasoning stated previously, Figure 10 displays overlaid Raman spectra from microtomed hair cross-sections collected at 0 and 240 min bleaching times. Immediately visualized is the increase in the 1040 cm−1 band intensity at 240 min bleaching. It should be noted that some background fluorescence in the 0 min bleached sample made the loss in the 509 cm−1 shift band a little less apparent however, integrating both band intensity Figure 10. Raman shift intensity spectra for 0 and 240 min bleaching times. Note the appearance of the 1044 cm−1 (-SO3−) peak after bleaching the fibers for 240 min. Figure 11. Raman scattering shifts as a function of bleaching time. Note the complimentary decrease in -S—S- (cystine) and increase in -SO3− (cysteic acid). Phe: phenylalanine (1003 cm−1).