A shampoo containing 10.5% SLE1S, 1.5% SLS, and 1.0% cocamidopropyl betaine sur- factants was formulated with histidine added at the 0.1% active level. Colored hair was washed for 12 cycles in tap water containing 0.06–0.09 μg/g copper with a sh ampoo with and without histidine and then analyzed for copper uptake using the ICP-OES method. Each wash cycle consisted of applying 0.1 g/g shampoo to the hair switch and lathering for 30 s, followed by a 30-s rinse repeated for a total of two shampoo applications. Ha ir was then dried in a hotbox at 80°C. This hair was then colored with two shades: an intense medium blonde/red (77/44) and a medium brown/violet (4/6). Each tint was mixed with a 20-volume hydrogen perox- ide developer (6%) in a 1:1 ratio and applied at a dose of 4 g of mix to each gram of hair. After 30 min, the hair was rinsed for 2 min, dried, and then color measured using a Minolta spectrophotometer. The hair was then washed with a shampoo containing 10.5% SLE1S, 1.5% SLS, and 1.0% cocamidopropyl betaine surfactants with and with- out 0.1% histidine for a total of 12 cycles with color fade measured using a Minolta spectrophotometer after 2, 5, 8, and 12 cycles. Each wash cycle consisted of applying 0.1 g/g shampoo to the hair switch and lathering for 30 s, followed by a 30-s rinse re- peated for a total of two shampoo applications. Hai r was then dried in a hotbox at 80°C. HDAP–AHT DYE SYNTHESIS 0.009 M HDAP and AHT were adde d to a 500-mL fl ask. Ethanol (20 mL) and am- monium hydroxide (0.083 M) were added to completely dissolve the solid. To an- other 250-mL Erlenmeyer fl ask, K3Fe(CN)6 (0.037 M) and water (100 mL) were added. A sonicator was used to dissolve the ferrate salt and form a homogeneous solu- tion. The ferrate solution was then added to the solution of dye precursors. The mix- ture was stirred continuously at room temperature for 30 min. The precipitate was fi lte red and washed with water to give a dark red solid (Z)-5-amino-4-((2-amino-1- (2-hydroxyethyl)-1H-pyrrol-3-yl)imino)-2-methylcyc lohexa-2,5-dien-1-one. This crude product was triturated with ethyl acetate and then dried in an oven (50°C) to afford 2.15 g of pure compound (yield 91.5% 98% purity confi rmed by electro- spray ionisation mass spectrometery and proton Nuclear Magnetic Resonance). DYE EXTRACTION 300 mL of metha nol was combine d wi th 150 mL of DI water in a 500-mL solvent bottle and allowed to equilibrate to room temperature before use. 100 mg of hair (four replicates fo r each copper level) were weighed into 4-mL vials, and 4 mL of extraction solvent was added to each vial and allowed to stand for 96 h. At the end of the extraction period, extracts were transferred to fresh vials and diluted 1:5 times and transferred to a disposable cuvette immediately before UV spectro- photometric analysis. A Cary 100 UV-vis spectrophotometer (Agilent, Santa Clara, CA) was used (waveleng th range: 400–800 nm, scan rate: 600 nm/min, and data inter- val: 1.0 nm). Absorbance data were normalized to account for differences in sample weight. REDOX METALS IN COLOR FORMATION IN A HAIR COLORANT 271

RESULTS AND DISCUSSION SOLUTIO N EXPERIMENTS The role of copper in accelera tion of color formation was tested fi rst in solution to simplify the sy stem to just dye precursors and hydrogen peroxide in the presence of varying levels of copper. Low levels of sodium sulfi te and ascorbic acid were added as reducing agents with the purpose of stabilizing the solution after making so that there was a reproducible starting solution color. Other precautions were taken to minimize air exposure before the reaction started, and tests were carried out to ensure vials were free of redox-metal con- taminants. Color formation as a function of time was measured with different copper levels for three dye couples—PPD–AHT, PPD– Res, and HDAP–AHT—after addition of hydrogen peroxide to the dye solutions with added copper levels. Color was measured from images taken of solution vials under controlled lighting. The three dyes showed dif- ferent ove rall rates of color formation, with HDAP–AHT being the fastest and PPD–Res being the slowest, but in all dye systems, copper addition increased the rate of color for- mation signifi cantly. Initial images were only taken after ~30 s after addition of hydrogen peroxide, and dye formation during this time was not measured. All dye solutions started with an L value of 99. Color formation values for L (lightness) for HDAP–AHT color formation show a dose response for increasing copper levels (Figure 1). Similar dose- response curves were obtained for PPD–Res and PPD–AHT. Change in L versus initial was calculated at 7 min for all dyes at each copper level and confi rmed a dose response of color formation acceleration versus copper level added for all three dye couples (Figure 2 ). Figure 1. L value change versus time for the HDAP–AHT dye couple. JOURNAL OF COSMETIC SCIENCE 272