CLEANING HAIR 313 injector and a flame ionization detector. The column is a Supelco 60 meter X 0.75 mm i.d. glass column coated with SPB-1 liquid phase to a film thickness of 1.0 microns. Detector temperature was 325øC. GC oven initial temperature was 220øC, held for eight minutes, ramping up to 310øC at 4øC per minute, and holding for 55 minutes. Figure 1 is a typical gas chromatogram of Spangler sebum we confirmed peak identifi- cations by mass spectrometry. Note that triglycerides are not detected under the column conditions used previous data indicate that these materials are easily removed by surfactants (7). We intend to modify our chromatographic system to test this con- clusion ourselves. RESULTS AND DISCUSSION The objective in this work was to determine if surfactants selectively remove sebum components from hair. Tresses were washed in dilute (0.01 to 0.1%) bulk (100 ml) detergent solution rather than attempting to simulate actual shampooing, because Thompson et al. (7) have shown similar results with improved precision by the bulk method. These low detergent concentrations are used to facilitate analysis of the sebum residues on the hair. If higher concentrations are employed, the recovery and subsequent analysis of the sebaceous residue is not practically feasible because of the very small amount of 456 20 21 22 23 24 25 Figure 1. Capillary gas chromatogram of Spangler sebum. 1. Tetradecanoic acid 2. hexadecanoic acid 3. n-eicosane (internal standard) 4. 9, 12-octadecadienoic acid 5. 9-octadecanoic acid 6. octadecanoic acid 7. n-docosane 8. n-tricosane 9. n-tetracosane 10. n-pentacosane 11. n-hexacosane 12. n-heptacosane 13. n-octacosane 14. squalene 15. n-nonacosane 16. hexadecyl dodecanoate 17. n-triacontane 18. n- hentriacontane 19. cholesterol 20. hexadecyl tetradecanoate 21. n-dotriacontane 22. n-tritriacontane 23. hexadvcyl hexadecanoate 24. octadecyl hexadecanoate and hexadecyl octadecanoate 25. higher molec- ular weight ester.

314 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS residue that is recovered. However, experiments in which soiled hair was handwashed using 10% surfactant, simulating actual use conditions, have shown results similar to those reported here for these test conditions (9). These handwashing tests are discussed later in the text. Five replicates for each of the following three surfactants have been performed: ß SODS- 1-- Sodium octeth- 1/deceth- 1 sulfate ß ALS--Ammonium lauryl sulfate ß SLES-2--Sodium laureth-2 sulfate The surfactants were chosen on the basis of total sebum removal data obtained using a wool substrate as a model keratin (9). Briefly, sebum removal from wool swatches is measured by monitoring the removal of a lipid-soluble dye (coadsorbed with the sebum) using a reflectance technique. These data (Table II) show that the surfactants may be considered as poor (SODS-1), medium-good (ALS), and good (SLES-2). Data are also shown for a second set of soil/wash conditions (B) (0.03 g soil/g hair 0.1% surfactant) values are averages of three replicates. As evidenced in Table II, values for the total sebum removed are in good agreement for the three experimental conditions (and two substrates) shown. For both A and B conditions, the amount of soil removed from hair is larger or equal to that for a wool substrate. The order of superiority of surfactants is also maintained (agreement with detergency theory), and the wool and hair values are close in magnitude. The total sebum removed under B conditions is larger than for A: as expected, the combination of lower soil loading and higher detergent concentration promotes better cleaning. The methods used in this work, soil/wash conditions and component identification, have been adapted from work reported in the literature (7). Three cleaning processes were described (7): bulk bath, finger squeeze, and controlled pressure/sponge the data show that the bulk bath method produces the most uniform results. Therefore, we have used the bulk method to provide as much precision in our experiments as possible and have drawn conclusions by statistical analysis of the data using a p value of 0.05 as the decision criterion. The conclusions in the Thompson paper (7) are based upon the less reproducible controlled pressure/sponge cleaning process. Thompson et al. (7) evaluated the shampoo detergency of three surfactants commonly used in shampoos: ALS, SLES-2, and AOS (sodium alpha olefin C14-C16 sulfonate) against fatty acids, cholesterol, paraffin waxes, wax esters, squalene, and triglycerides. The gc system used in our work did not allow for detection of the triglycerides and all other fractions at the same resolution. The triglycerides have much longer peak reten- Table II Data Comparing Total % Sebum Removed for One Soil/Wash Cycle From Hair and Wool Surfaces % Removed % Removed % Removed Surfactant (Cond. A) (Cond. B) (Wool) SODS-1 40.7 + 15 56.7 + 25 35 -+ 4 ALS 72.4 -+ 9 97.3 + 2 79 +- 3 SLES-2 93.7 + 3 97.6 -+ 2 88 -+ 2 Condition A: Hair soiled at 0.04-0.055 g soil/g hair and washed with 0.01% surfactant. Condition B: Hair soiled at 0.03 g soil/g hair and washed with 0.1% surfactant. Wool: 3-inch X 4.5-inch wool challis swatch soiled with sebum/lipid-soluble dye soil.