THE GREEN HAIR PROBLEM 5 Unoxidized (From Treatment Bath #1) Oxidized (From Treatment Bath #1) % . Unoxidized Conditioner Pretreated (From Treatment Bath #2) Oxidized Conditioner Pretreated (From Treatment Bath #2) Figure 1. EDXA-SEM photographs of cross sections of the hair fibers treated with the copper sulfate containing simulated swimming pools at 6 pH. Effects of mild oxidation and conditioner pretreatment are evident, and they correspond with the results of Table III. explained in the light of the above hypothses. Our own amino acid data on mildly oxidized hair such as that used in the present experiment and some of the relevant literature (16) show that such oxidation causes little changes in carboxyl and amino group content. Hence these sites cannot be readily accepted as the copper binding

6 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS sites. Oxidation, even at mild levels, does decrease the half-cystine content and increase cysteic acid, and therefore the sulfonate groups perhaps participate, directly or indirectly, in copper absorption. EFFECT OF CONDITIONER PRETREATMENT The inhibition of copper uptake caused by the quaternary ammonium compound conditioner pretreatment is not surprising. The cationic active ingredient is likely to tie up the anionic metal-binding sites in hair, preventing copper absorption. The affinity of the conditioner to hair may be established from the results of another experiment conducted in our laboratory. The experiment involved a repeated process of immer- sion of hair tresses in a dilute aqueous solution of the conditioner, drying and subsequent treatment with a simulated swimming pool for an hour. Soxhlet extraction of the hair tresses with a 50/50 mixture of chloroform and methanol after 25 cycles of treatment yielded a residue amounting to about 5% by weight. This residue and the "sticky" feel of the hair before extraction suggested that the conditioner ingredients deposited on the hair to provide a coating which survived the swimming pool exposure in each cycle of treatment. ATTEMPT TO REMOVE THE GREEN COLOR Separate portions of the green colored unoxidized and oxidized samples (from Treatment Bath #1) were washed with a leading baby shampoo, a balsam and protein shampoo, or a strong aqueous solution of sodium lauryl sulfate. None of these washings removed the green color. Other separate parts of the green tresses were treated with ethylene diamine tetraacetic acid (EDTA) at 10 pH, dilute ammonium hydroxide or dilute nitric acid. The EDTA treatment imparted a golden-brown color, the NH4OH treatment gave a silver gray color and the HNO 3 treatment left the green color unchanged. Thus no immediate method of removing the green color appeared evident. ROLE OF CHLORINE IN COLORATION In addition to the heretofore described experiment, another experiment was carried out by immersing water or conditioner-pretreated blond tresses in separate beakers containing about 12 ppm copper (in the form of CuSO4 ß 5H20) in distilled water adjusted to 3.8 pH using HC1. After several hours, the conditioner-pretreated hair turned very light green and the water-pretreated hair turned light green, suggesting that chlorine is not necessary for generation of the green color. ANALYSIS OF A REAL-LIFE CASE A sample of '•green" hair collected from a young, female, regular swimmer with long blond hair which had developed streaks of green color was analyzed for copper content. Atomic absorption analysis showed that the "green" hair sample contained about 4220/ag of copper/g of hair. This value compares with the copper content of the oxidized hair treated with the copper sulfate solution at 6 pH for ! hr in our experiments (Table III). The "green" and "normal" hair strands from the same subject