CATIONIC POLYMER ADSORPTION 91 treatments ranged from 10 to 40 mg polymer/g hair, depending on solution concen- tration. Containers were tightly covered and placed in a 40øC shaker bath for a treatment time standardized to 45 minutes, unless otherwise noted. At the end of the treatment time, solutions were decanted for analysis by titration. DETERMINATION OF CATIONIC POLYMER LEVEL ON HAIR Changes in cationic concentration following immersion of hair tresses were determined by colloid titration. Cationic polymer uptake is reported in mg cationic polymer/g hair: where Uptakecationic potymer(mg/g hair) = (mginim t - mgfin•t)/g hair [Eq. 4] mginiti• t = mg cationic polymer in solution before treatment and mgnn•t = mg cationic polymer in solution after hair treatment. RESULTS AND DISCUSSION The colloid titration method is based on a color change in the o-Toluidine blue indicator with the first excess of KPVS titrant. An anionic polymer, KPVS, is added to the analyte solution where the cationic polymer and anionic KPVS form a preferential complex. The cationic indicator remains unreacted. When all cationic analyte has complexed, the excess titrant begins to interact with the o-Toluidine blue indicator, resulting in a color change from blue to violet. The equivalence point of the titration is obtained from the titration curve and used to calculate the amount of polymer present in solution by Equation 3. Polymer uptake by the hair is inferred and calculated from differences in solution concentration before and after hair treatment by Equation 4. The colloid titration method is subject to interferences (2) including: 1) indeterminate stoichiometry between anionic titrant and cationic analyte, 2) time effects in determin- ing the endpoint, and 3) ionic contaminants including leached hair proteins and pig- ments. These difficulties are addressed in the present study. The relationship between the Polyquaternium-10 and the KPVS titrant was tested for linearity. Eight repetitions of the titration demonstrated stoichiometry of the relation- ship within the concentration range of interest (Figure 2). Pooled standard deviations of 0.05 and 0.01 are noted for the slope and intercept, respectively, of the averaged determinations. The problem of endpoint time effects is acknowledged and normalized by titrating test solutions at contant KPVS delivery rate. This minimizes kinetic effects that govern the approach toward equilibrium of colloid complexation and indicator response. In addi- tion, the colorimeter equipped with a sensor electrode is capable of an immediate response and resolves any difficulties in visual detection of titration endpoints that are often subtle. Ionic contamination was not a factor in this work, as all cationic polymer solutions were kept free of surfactants or other interfering compounds.

92 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Finally, the leaching of soluble protein and pigment from bleached hair into the cationic polymer solutions was examined. Detrimental effects of artifacts arising from these constituents were minimized by soaking bleached hair thoroughly. This was demon- strated by noting that the endpoint of the buffer-indicator blank remained unchanged from a blank that also contained the thoroughly rinsed, bleached hair. Previous colloid titration work to quantify cationic adsorption onto a negatively charged substrate was performed by reference to a calibration curve (5,10-12). In the present study, the equivalent weight of the polymer was a constant value within the concen- tration range of interest. This allowed accurate and reproducible determination of the amount of cationic polymer in solution. Several different concentrations of Polyquaternium-10 ranging from 0.05 to 0.20% w/v were evaluated. Aliquots of these solutions were titrated and the equivalent weight of Polyquaternium-10 was found from the slopes of the individual regression lines. Lin- earity of response between KPVS and Polyquaternium-10 was evident within the con- centration range studied (Figure 2). The equivalent weight of Polyquaternium-10 was found to be 877.7 + 19.1 g/equiv by application of Equations 1 and 2 and the averaged slope (2.2 ml/mg) of the regression lines (Figure 2). This result correlated very closely with the 875 g/equiv value calculated independently from direct and standard substit- uent analysis of the polymer that is, analysis for hydroxyethyl molar substitution (MS) by the Zeisel method and determination of nitrogen content (Kjeldahl method) yielded values of 2.24 and 1.45 percent, respectively. Assuming that the cationic moiety is the (2-hydroxypropyl)trimethylammonium chloride (CH2CHOHCH2N + (CH3)3C 1 -) res- idue, a nitrogen level of 1.45 percent yields a degree of substitution (DS) of 0.36. These MS and DS values yield a molecular weight (MW) of 315 grams per polymeric repeat (monomer) unit from which an equivalent weight of 875 g/equiv, i.e., MW/DS, is obtained. The close agreement between these two independent analyses suggests a one-to-one stoichiometric relationship in charge neutralization between the KPVS ti- trant and Polyquaternium-10. Difficulties in determining the equivalent weight of Polyquaternium-10 are worth mentioning. The use of a sodium citrate buffer (pH 6.0) to prepare the polymer solution gave variable results, especially when Polyquaternium-10 solutions were examined over a period of several days. This variability was attributed to ion exchange between the relatively large citrate ion and the chloride of the Polyquaternium-10. However, the formation of soluble complexes of Polyquaternium-10 with the tribasic citric acid may also be a source of stoichiometric variability. When the solvent system was changed to phosphate buffer, the equivalent weight of the cationic polymer remained constant when examined for up to five days after preparation. In order to calculate the equivalent weight of Polyquaternium-10 accurately, the rela- tionship between the cationic polymer and KPVS must remain constant. However, the relationship between KPVS titrant and Polyquaternium-10 deviated from linearity when the amount of Polyquaternium-10 present exceeded 400 •xg. A similar problem was noted by Onabe, who also used colloid titration analysis (12). The deviation was attributed to an excess in cationic charge on the cationic polymer-KPVS colloidal product at high levels of cationic polymer. Residual cationic charge hindered selective adsorption of KPVS on the 0-Toluidine blue indicator and changed the titration end-