16 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 4O o.. , 30 lo Figure 3. Comparison of scale lifting phenomena in root and tip sections of hair fibers. A: Unextended hair fiber. B: Random scale lifting. C: Common scale lifting. D: Extreme scale lifting. E: Hair fiber breakage. from the scale face below and the formation of stress concentrations at the scale edges should also be considered as a source of increased fluorescence intensity. The random scale lifting and brilliant scale lines become more common upon further extension, and finally scale lifting becomes extensive in frequency and angle (Figure 1, right), followed shortly thereafter by breaking of the hair fiber. The extension levels at which specific scale lifting phenomena occur were recorded. SCANNING ELECTRON MICROSCOPY The extended fibers were mounted on metal stubs and were studied in a JEOL-JSM-12 and more recently in a Hitachi S-4500 field emission scanning electron microscope. REVERSIBILITY OF EXTENSION Fibers were extended to 30-35% under ambient conditions, held in the extended state for -5 rain, and then released and transferred into water at room temperature. After 20 hours the fibers were dried and conditioned (65% RH, 2 IøC) for at least 24 hours. These fibers were then re-extended under the microscope, and the scale lifting phenomena were re-examined. The mechanical properties of fibers with and without prior 30% extension were determined in water by using an Instron tensile tester. RESULTS AND DISCUSSION In the initial studies, root sections of 18-in-long brown European hair (from DeMeo Brothers, New York) were used, in order to eliminate some of the damage that is inflicted on the cuticular region by standard grooming practices and to ensure the

STRESS RELEASE IN HAIR CUTICLE 17 presence of multiple cuticle layers in the cuticular sheath. In our efforts to quantify the phenomena observed during extension, we decided to establish the extension level at which characteristic scale lifting phenomena are observed. We differentiate among (A) the unextended hair, (B) the first randomly distributed lifting of individual scales, (C) common scale lifting along the whole length of the fiber, (D) extreme scale lifting both in frequency and angle, and finally, (E) breakage of the fiber. The results for the extension of the root section of ten (10) hair fibers are shown in Figure 2. When the studies were extended to the near tip section of 18-in-long hair, we observed that the onset of scale lifting occurred at significantly lower extension levels, with the other stress phenomena shifted similarly to lower extension levels (Figure 3). Only the final breakage of tip and root sections occurred at approximately the same level of extension. This shift to lower extension levels clearly indicates that grooming and exposure to stresses experienced during the longer lifetime of the tip section have caused a loosening of scale edges, possibly involving damage to the endocuticular layer as well as damage to and conceivably loss of the intercellular domains, which is definitively irreversible in nature. In an effort to confirm the cuticular damage observed in auto fluorescence and to explore details of failure at the scale edge resulting in the scale lifting as well as any other potential events occurring on the surface cuticle during extension, the extended fibers were transferred to the scanning electron microscope. Highlights of the SEM study displaying the various damage phenomena at specific extension levels are seen in the micrographs shown in Figures 4a-1. The most dramatic damage phenomenon is the scale edge lifting, which increases in frequency and angle with increasing levels of extension. The frequent appearance of granular material underneath lifted, chipped away, or broken-off surface cuticles, which has been associated with endocuticular debris by Swift (9), suggests that the endocu- ticular layer is indeed a region of weakness within the cuticle cell. Based on these observations, we can conclude that endocuticular failure precedes scale edge lifting. As pointed out above, the difference in amino acid composition of the different layers of the cuticle cell leads to large differences in swelling and, with it, deformability. During extension of the fiber, this difference in deformability sets up shear forces between these layers, as indicated schematically in Figure 5. At higher extension levels, this shearing action can lead to stress concentrations and finally failure at the edge of the endocuticle. This edge failure results in a partial "delamination" within the scale structure, and under the influence of the shear forces the upper layers of the surface scale are lifted up, starting at the scale edge. J. A. Swift (9) observed the phenomenon of endocuticular failure during wet combing, where failure of the swollen endocuticular layer within the cuticle cell resulted in severe scale fracturing. It is well known that the extension of keratin fibers, at least within the yield region up to extensions of 25-30%, is totally reversible upon release of the fiber and its immersion in water. Similarly, extensive and, for some properties, complete recovery of mechanical properties has been observed. In our study, we have observed a total length recovery after release and water immersion and a partial recovery of the extension-induced cuticular scale lifting described above. This recovery was similar in tip and root sections, with most lifted cuticles returning to their original apparently tightly stacked configuration (Figure 6). However, occasional scales that had been lifted to an extreme angle remained