ANIONIC SURFACTANT RINSABILITY 85 Table IVb Delta Absorbance Values Showing Absence of Fluorescein Deposition (485 nm) From Bar Slurries Product 280 nm 485 nm Bar A 0 0.01 Bar B 0.05 0.02 Bar C 0 0.01 Experimental conditions: Porcine skin, 15-sec lather treatment, rinse with 100 ml water (260-ppm hardness), extraction with 2 x 10 ml 80:20 MeOH:water. Delta absorbance is the difference in absorbance between methanol:water extracts of skin treated with bar slurries with and without fluorescein. Values 0.0• are within experimental error. DISCUSSION A key finding of the present study was that the highest level of fluorescein retained on skin occurred from a dispersion of fluorescein in water. As already discussed, in vivo treatments of forearms with a 0.005% (50 ppm) dispersion of fluorescein also exhibited significant retention of the dye after mild rinsing. Importantly, the extent of deposition in in vivo studies was found to be less at pH 9.2 compared to pH 7. These results indicate that fluorescein can indeed be retained on the skin but totally independent of surfactant residue. The extent of deposition is a function of the pH of the aqueous solution. These results do not agree with those described in reference 2, where it was reported that a "1% fluorescein dye solution left no detectable residue on the skin after 15 minutes" (rinsing conditions unspecified). In contrast, our attempts to repeat the 1% fluorescein experi- ments on porcine skin showed significant dye deposition at pH 5 and measurable deposition even at pH 8. The reasons for this discrepancy are unclear, but it again points to the criticality of differences in pH and even small differences in experimental pro- cedures. It is thus possible that the 1% "solution" used by Wortzman et al. (2) was a neutralized sample and had a high pH. It is noteworthy that fluorescein is soluble in distilled water only to a level of about 0.002% (20 ppm) unless it is neutralized with base (see Figures 5-7). The pH study dearly shows that the extent of fluorescein deposition is inversely related to its solubility in the aqueous phase and actually de- creases in the presence of surfactants, especially those that bind strongly to skin. Since the deposition behavior of fluorescein is strongly dependent on solution pH and the type of counterion (for example TEA vs Na because of the differences in their ability to buffer the pH), it can be concluded that fluorescein is not a suitable probe to report the behavior of solutions/bar slurries having different pH values and counterions. The second key finding is that fluorescein has essentially no tendency to associate with anionic surfactants that are present either in miceliar form or that are strongly bound to the stratum corneum. Although we tried to mimic the very gentle rinsing conditions employed in reference 2, we were unable to detect any residual bound fluorescein in vitro from any of the surfactant systems studied, either visually or by solvent extraction. Although the solvent extraction procedure employed in reference 2 can give a false positive for fluorescein because of extraction of skin components that adsorb at 280 nm, the use of the most sensitive 485 peak also failed to show any residue. Thus, it is concluded that the fluorescein rinsing assay does not measure any intrinsic interaction between anionic surfactants and skin since many of the surfactant systems studied here have been previously shown to leave significant residual surfactant strongly

86 JOURNAL OF COSMETIC SCIENCE bound to the keratin proteins of the stratum corneum, e.g., pure sodium dodecyl sulfate (3,21), sodium laurate, Bar B (high glycerin, TEA-based soap), and Bar C (pure soap) (1). Therefore, whatever values were reported in reference 2 must have been highly sensitive to the detailed experimental procedures employed. We will return to this point shortly. We are now in a better position to understand what the fluorescein rinsing assay actually measures and propose why it does not correlate with measurements of the actual binding of surfactants to skin and the damage that they can produce. When slurries containing fluorescein are applied to and rinsed from the surface of the skin, fluorescein will track changes that take place in the aqueous phase of the surfactant mixture. Since fluorescein does not partition into surfactant micelies or other association structures, it cannot report on their location. If a limited amount of cool water is used to rinse the slurry under minimal agitation, a finely dispersed precipitate of fluorescein can get trapped in the crevices and folds in the skin surface. Since TEA soaps have high pH and fluorescein is highly soluble in these solutions (see Figures 6-8), the bulk of the soap solution will have a tendency to be rinsed off more quickly from a surface in this type of test, even though the surfactant has a considerable molecular interaction with skin and is strongly bound to the corneum proteins on a molecular scale. Any differences between formu- lations are driven primarily by pH, buffer strength, and type of base rather than sur- factant residue, i.e., the test measures fluorescein residue, not surfactant residue. Even though the differences reported in reference 2 are an artifact of the probe used, it is appropriate to comment on the rinsing conditions employed. Although exaggerated tests are often very useful in predicting product performance attributes, e.g., lather (22) and mildness (19), we believe that the test conditions are very unrealistic and can provide a misleading measure of the rinsability of cleansing compositions in everyday use. The procedure described in reference 2 employs the minimum volume of cool water ("room temperature") that is required "not to leave a visible residue on the skin" and is performed without rubbing or any agitation. These conditions are very different from those commonly employed by consumers in the cleansing process. For example, habits studies (23) indicate that an average U.S. shower lasts approximately 4-5 minutes, with water at a temperature of about 32-43øC, pumped at a rate of about 6.5 liters per minute. Similarly, an average face wash lasts about 30-45 sec, with the majority of people using water at a temperature in the range of between 32øC and 43øC and rarely below 27øC. Although the amount of water used, and its flow rate, are far below the values quoted for showering, considerable rubbing take place during the rinse. The above analysis indicates that a more realistic measure of rinsing might, in fact, be the surfactant residue that is strongly bound to the skin (e.g., stratum corneum) since it resists even the vigorous rinsing that is commonly found in practice. If this measure of residue were employed, then Bar A has a significantly lower surfactant residue than either Bar B or Bar C, as has previously been shown (3). Spectroscopic studies suggesting this conclusion will be published in the future (24). SUMMARY AND CONCLUSIONS Fluorescein does not track the binding of surfactants to skin and, thus, cannot measure intrinsic interactions between a cleansing composition and skin. Since fluorescein does