PHOTOLIGHTENING MECHANISM OF RED HAIR (%) 90 80 70 60 50 40 30 non­ irradiated a UV VIS (%) 90 80 70 60 50 40 30 non­ irradiated b UV Figure 5. Percentage of disappeared melanin granules. a: Red hair. b: Blond hair. 53 VIS around 26% from that of the non-irradiated sample, while VIS-light irradiation resulted in an increase of only 5%. In the case of blond hair, the increased percentage was around 10% after either UV- or VIS-light irradiation. Therefore we can say that the melanin in red hair is much more sensitive to UV light than to VIS light. Its UV sensitivity is also much higher than the sensitivity of blond hair to either light. The color change of red hair was similar in either case for UV- or VIS-light irradiation (Figure 2). By comparing this with results from the irradiated cross section, it is proven that the effect of UV light decreased when irradiated on hair. Damage to hair structure. Examples of SEM images for the cross sections rinsed with water after irradiation are shown in Figure 6. Many cracks, up to micrometers in length, formed on the UV-light-irradiated samples (Figure 66). In contrast, no cracks were observed on the VIS-light-irradiated cross sections (Figure 6a). Therefore it was ascer­ tained that UV light damages red hair tissues, while VIS light causes no recognizable damage. These findings indicate that the following hypothesis previously stated for blond hair (1) applies equally to red hair: In hair, UV light is absorbed by hair protein and is attenuated prior to reaching the melanin granules. Because of this, the decom position of melanin granules inside hair fibers is lower than the decomposition of melanin granules that are exposed on hair cross sections that directly receive UV light. On the other hand, VIS light reaches melanin granules inside hair without being attenuated. As a result, the degree of lightening in red hair caused by UV light is reduced, even though the melanin of red hair was originally believed to be much more sensitive to UV light than to VIS light. In this study it was found to be at the same level as that of VIS light. In consideration of the types of melanin in blond and red hair, these results indicate that pheomelanin is much more sensitive than eumelanin to UV light. It appears that when

54 JOURNAL OF COSMETIC SCIENCE a b Figure 6. Cross section of red hair (after irradiation and washing). a: After VIS-light irradiation. b: After UV-light irradiation. UV light was irradiated on hair under these experimental conditions, the pheomelanin began to decompose, which caused the red hair to lighten. On the other hand, eumelanin was hardly affected by the UV light irradiation and the blond hair did not therefore lighten. In the case of VIS-light irradiation, the degradation of melanin in blond hair was found to be a little higher than that of red hair however the difference was within the bounds of experimental error (Figure 5 ). From this, it is supposed that the two types of melanin are similarly sensitive to VIS light, or perhaps eumelanin is just a little more sensitive than pheomelanin. PHOTOLIGHTENING OF ISOLATED GRANULES FROM RED HAIR The isolated granules obtained from red hair are judged from their appearance to be pheomelanin however, they are possibly denatured through the papain treatment. The photolightening behavior of the isolated granules was also studied and compared with the results of the irradiated cross section. The photolightening behavior of the granules is shown in Figure 7. The degree of photolightening was a little higher for UV light than for VIS light. However, the difference is so small that we cannot say this is consistent with the much clearer result of the irradiated cross section, where the melanin remained chemically intact. This clarifies that pheomelanin, denatured by treatment with papain, is not suitable for the study of photosensitivity. CONCLUSIONS Both VIS and UV light lighten red hair. Unlike blond hair, washing after irradiation is