32 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS oxygen as oxidant. The Piedmont hair and wool gabardine were commercial samples from Arkinco Inc., New York, and Test Fabrics Inc., Middlesex, NJ, respectively. Hair was fabricated into 2.5 g swatches secured by wax, and the wool cloth was cut into 2" X 2" squares weighing 0.6-0.7 g. After soaking in the formulation, the dyed material was rinsed with deionized water for 5 sec and then dried. The amount of dye deposited on the fiber was determined from the reflectance spectrum measured by a Hunter LabScan Spectrocolorimeter, Model LS-5100. In general, the reflectance spectra were similar to the transmission spectra of the dyes in solution but with broader, less well-defined peaks. Spectra from wool were sharper than those from hair. RESULTS EXPERIMENTAL DESIGN We have used a standard response surface experimental design. Of the variety of avail- able designs, the 5-variable, central composite rotatable second-order design (4) ap- peared to offer the most information from the least number of experiments in our system. The specific factors studied were: © dye concentration--[PPD] or [BHP] varied from 1.0-7.0 mM/L, with equimolar coupler added © pH--9 to 11, adjusted with NH4OH/NH4C1 solution ß H20 2 concentration--1.0 to 11.0% (0.3 to 3.2 M) © surfactant concentration--Nonoxynol-9 varied from 0.5 to 4.5 wt % © dyeing time--5 to 25 minutes However, we also evaluated ammonia concentration over the range 0.5-4.5% at fixed pH, but it had negligible effect on dye deposition, and sodium chloride concentration, where a significant salt effect was observed. All experiments were, therefore, carried out at constant ionic strength (1.1) by adding NaC1. The design required 32 dyeings for each substrate using the values of each of the variables specified in Table I. After dyeing, the reflectance of each sample was measured and these values were con- verted into a response surface using the method described in reference 4. The specifics of the design and the data manipulation will not be described here, since they were carried out precisely as described. Table I Values of Variables Used in Experimental Design [Dye] pH Time [8202] [Surfactants] 1.0 mM/L 9.0 5 min. 1.0% 0.5 wt % 2.5 9.5 10 3.5 1.5 4.0 10.0 15 6.0 2.5 5.5 10.5 20 8.5 3.5 7.0 11.0 25 11.0 4.5

OXIDATIVE DYEING OF KERATIN 33 REFLECTANCE MEASUREMENTS There is no absolute mathematical relationship between the measured reflectance (R) of a dyed fiber and the concentration of adsorbed dye (C). However, the Kubelka-Munk Equation [ 1] is an approximation which is commonly used because of its ease of practical application (5). (1 - R)2/2R = KC ........... [1] In this equation, K is a constant for each dye, and its value was determined using fibers dyed with pre-formed indo dyes. Hair and cloth swatches were separately dyed with solutions of various concentrations of the indo dyes, and the amount of dye removed from solution by the fibers was measured spectrophotometrically. By mea- surements of R at various solution concentration differences (i.e., absorbed dye intens- ities), K can be evaluated graphically (Figure 1). The observed linearity shows that K is generally invariant with concentration as predicted by the equation and has the values for the PPD/AHT dye on wool 101, on hair 571 and for the BHP/NAP dye on wool 107 and hair 333. These values were then used to evaluate adsorbed dye concentrations from the reflectance readings of dyeouts from reactions of the individual couples. DISCUSSION Dye can be deposited in a fiber by any of the following means: :,7• n :-• I I 6 1 4,4, "-' 0 9 • r-Pt [v!,3 le: 5' F:',,.,'*• •,.'{ •:O r" E:: • ::.'J ,' C-r'r-, ':,:':/:-':,-• l •_. _.,. ,_. _ .... ._.._. Figure 1. Variation of reflectance function with concentration of adsorbed dye for dyeing wool fabric with 2-amino-5-methylindoaniline (PPD-AHT dye). Slope of line is Kubelka-Munk constant K.