EFFECT OF UV RADIATION ON HAIR STRUCTURE 107 All of what has been written so far indicates that oil-soluble sunscreens present some challenges from the formulation point of view. The formulations need to be cosmetical­ ly acceptable, the sunscreen needs to be delivered to the hair in a more or less uniform manner, and it should be deposited onto the hair and retained by the substrate to pro­ vide protection for a period of hours or maybe a couple of days (until reapplication). Also, as has been reported so far, there is little learning that is transferable from one formulation to another all the data reviewed appears to be very formulation­ dependent. In summary, traditional water-insoluble UV absorbers are difficult to deliver to the hair, and traditional water-soluble UV absorbers can be easily washed off from the hair during the rinsing steps of hair care product application. The term traditional is used here to describe sunscreens that have been mainly used for skin protection and sun care formulations and that have also been investigated for the protection of human hair. A different class of sunscreens, cationically modified, is discussed in the next paragraphs. Zulli (1 7) describes the delivery of lipophilic UV absorbers (isoamyl p-methoxy­ cinnamate and avobenzone) to the hair via preparation of cationically charged nano­ vesicles. The nanoparticles described were prepared by high-pressure homogenization of a phospholipid dispersion. With this mechanism the particles can be designed to be positively charged or even negatively charged. The positively charged particles, unlike the negatively charged ones, show affinity for the hair, even from rinse-off treatments. A different class of UV filters was investigated by Gao and Bedell (18). In this case, affinity was achieved by cationically modifying the UV-absorbing molecule. The authors investigated the effectiveness of cinnamidopropyltrimonium chloride (CATC, a cationi­ cally modified UV absorber) as it compared to octylmethoxycinnamate (OMC) on natu­ ral grey hair. The paper describes a shampoo formulation based on sodium lauryl sulphate (10%) and 2% sunscreen (CATC or OMC). Under these conditions, OMC, an oil-soluble com­ pound, was solubilized by the surfactant. CATC was thought to form a complex with the surfactant, subsequently to be solubilized by it. The application of the formulation to the hair was by immersion and it was followed by a rinsing step. The application (and rinse) was performed after every 24 hours of irradiation the total irradiation time was 20 days (27°C and 65% RH). The distribution of the UV irradiation was given by combining the following two sources (no VIS light source was added): 280-320 nm (UVB), 0.14 mW cm- 2 (= 1.4 W m- 2 ), and 320-4 00 nm (UVA), 0.4 9 mW cm- 2 (= 4.9 W m- 2 ). The irradiated and non-irradiated hair was analyzed for color change (CIE L*a*b*, DL, Db, and DYI, yellowing index), fiber diameter and cross-sectional area, tensile strength, wet combing force, dynamic contact angle, and transverse swelling. The substantivity (19) of CATC, under the conditions employed, was found to be 4.5 g of material deposited per 100 g of hair. Only trace amounts of OMC deposited under the same conditions this is not surprising since the formulation was not optimized to deposit OMC. On the other hand, CATC, being cationically charged, has an intrinsic affinity for the hair fiber. Although the paper does not go into detail on the interactions between the cationically modified UV absorber and the anionic surfactant, it is plausible to expect that the two oppositely charged moieties will form some sort of complex, and perhaps it is the complex that is depositing onto the hair as well as, or instead of, the free cationic UV absorber molecule.

108 JOURNAL OF COSMETIC SCIENCE The following points summarize the results presented in this paper: 1. Gray hair changes color, turning yellow upon exposure to UV radiation. CATC effectively prevents gray hair from changing color when compared to the untreated samples. 2. Brown hair treated with CATC changes less in color (photobleaching) than the untreated, UV-exposed sample. 3. The OMC-treated hair was substantially less protected from UV radiation, with results more similar to the untreated hair than to the CATC-treated hair (OMC deposited on the hair in much lower concentration than CATC from the test for­ mulation studied). The test hair samples were analyzed for changes in color, wet combing force, tryptophan concentration, cuticle abrasion, and change in contact angle. 4. When the treated hair tresses were tested for tensile strength (Young's modulus and stress-to-break), all four hair tresses tested (not exposed to UV, CATC + SLS, OMC + SLS, and SLS alone) scored with approximately the same difference, from one to the other. The most damaged was the SLS-treated hair, and the least was the unexposed CATC + SLS gave more protection than OMC + SLS. A similar trend was observed when transverse swelling was measured. 5. Hair tresses pretreated with SLS + CATC, and irradiated with UV, showed improved wet compatibility when compared to the unirradiated samples. 6. Gray hair is more sensitive to UV-induced damage than dark brown hair. This study investigates the properties of two UV absorbers that have different solubility and intrinsic charges it is not surprising that OMC performed quite differently from CATC in the tests described. The mechanism of deposition of the UV absorbers onto the hair was not described by the paper, but it is reasonable to assume that the mechanism would involve both hydrophobic forces (OMC, CATC) as well as ionic interactions (CATC). It is interesting that although OMC was solubilized by the surfactant system, it did show some degree of efficacy in some of the experiments. A totally different approach to photoprotection was investigated by Pande et al. (20). They investigated the role of hair dyes (oxidation, as well as direct dyes for hair color) in the photoprotection of human hair (as monitored by tensile strength measurements). The total irradiation time was between 96 hr and nine days (13.9 kJ cm- 2 , 31.2 kJ cm - 2 ), depending on the dye product investigated. The instrument used was an Atlas Fadeometer Ci35A, 50°C, 50% RH. The data presented in this article showed that natural unpigmented hair (Piedmont) is more damaged by irradiation than naturally pigmented hair (brown). In addition, hair color (artificial) does provide protection from UV-VIS irradiation in all cases, the darker the color shade, the more the protection. A distinct difference was also identified between oxidation hair color and direct hair color. The testing protocol used involved subsequent cycles of hair dying and shampooing, which is very appropriate when studying the fading behavior of artificially colored hair, even if the dyes used are oxidation colors (oxidation dyes are usually considered to be less prone to fading by shampoo however, red shades are more sensitive to it than dark shades). Another aspect to keep in mind is that while oxidation colors have the ability to penetrate completely though the hair fiber, direct dyes do not always penetrate much into the fiber. Depending on the depth of the shade, this may have a substantial impact on the results of irradiation. A direct dye that has heavily deposited on the surface of the hair fiber may act as an effective screen for any radiation to penetrate into the fiber, thereby protecting it.