310 JOURNAL OF COSMETIC SCIENCE photodamage is proportional to the square root of the amount of free-radical initiator present in the fiber. The mechanism assumed that the S-S group was the initiator, giving two -S' free radicals by homolytic scission. Leroy et al. (8) and Dubief (9) investigated the role of moisture in the photobleaching of hair. Their research was limited to extreme RH levels, namely 5% and 88% only. Their data showed an increase in bleaching, measured as luminance, with exposure time, as shown in Figure 1. They also noted a greater loss in tensile properties, measured under deionized water, when hair was irradiated in the presence of moisture as per the 1 stretching index shown in Figure 2. Also, a radiation dose of 105 J/cm 2 approximately equals an exposure period of 360 hours. Based on the concepts suggested by Wei, we hypothesize that moisture, in combination with trace minerals such as iron, is a source of free radicals in the presence of UV radiation. This leads to the notion that an increase in photodegradation will occur with increasing amounts of moisture during irradiation, up to a certain concentration. This work tests the hypothesis by investigating the role of moisture in the photolysis of melanin-free human hair exposed to simulated solar radiation. Our objective is to characterize the damage in hair keratin as a function of RH during exposure to simulated solar radiation in air. The result is a better understanding of the photodegradation processes of hair keratin. Such knowledge may aid in formulating hair care products that minimize photolytic damage. MATERIAL Bleached hair was avoided because the bleaching process damages the fiber (10). There- fore approximately 6-in-long root ends of 18-in-long virgin, unaltered, naturally mela- nin-free, Piedmont hair samples, supplied by DeMeo Bros., New York, were used. WEATHERING About 120 individual Piedmont hair fibers (6-in-long) were mounted in a single layer 25 - - •, - .XENOTEST (88%) _-- SUNTEST (5%) O t"= ,,, 20 0 50000 100000 150000 Doses (Jlcrn 2) Figure 1. Bleaching of pigmented hair exposed at two RH levels (8).

PHOTODEGRADATION OF HUMAN HAIR 311 x 30 .-- 20 --,e,--XENOTEST (88%) • SUNTEST (5%) ...e ß ß ß ß E 10 ''''' '• ,•' ' ½ 0 0 50000 100000 150000 Doses (J/cm 2) Fiõur½ 2. Ch•es i• [•-st•etchJ• j•dex o[ pigmented h•J• exposed •t two EH levels (8). parallel to one another on plastic frames using double-sided tape. The mounted speci- mens were exposed to simulated solar radiation in the Atlas weatherometer at RH levels of 10, 20, 30, 50, and 70% for 100, 200, and 300 hours at 50øC. The Atlas weatherometer uses a Xenon arc lamp to simulate the solar spectrum. Boro- silicate filters were used to provide an "average" 45 ø Miami sunlight. The energy density at 340 nm was maintained at 0.3 W/m 2, resulting in a total energy density of 41.27 mW/cm 2. It should be noted that only about 10% of this energy was in the UV region and that the remaining energy was in the visible region of the solar spectrum. Automatic controllers maintained set temperature and humidity levels in the chamber (11). TESTING Three tests were employed to assess photodamage: 1. The difference in tensile properties between untreated and weathered hair was mea- sured with the fibers totally immersed in deionized water. 2. Swelling tests were conducted on hair weathered for 300 hours, with an eye towards assessing changes in crosslink density as a function of RH during weathering. 3. The Fourier transform infrared/attenuated total reflection (FTIR/ATR) technique was used to assess surface damage of hair weathered for 300 hours at various RH levels. TENSILE TESTS A laser-scan micrometer supplied by Diastron Ltd., UK (Mitutoyo, model LS 3100) was used to measure the fiber cross-sectional area. The instrument employs a 1.0 mW 670- nm wavelength laser. The sample is placed in air, obstructing the laser beam, and rotated slowly by a motor. The resultant shadow falls on a sensor. The micrometer assimilates the data from the sensor and continuously measures the fiber diameter. The instrument has a precision of about 5 pm. The included LSMVB software analyzes the data from the micrometer, and identifies the minor and major axes of the fiber. From this data, the