234 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS EXPERIMENTAL MATERIALS The cationic cellulosic polymers used were UCARE © Polymer JR 400 (CTFA designa- tion Polyquaternium-10) and QUATRISOFT • LM-200 (CTFA designation Polyqua- ternium-24), both products of Union Carbide Corporation. Three additional cationic polymers were also examined, Hercules RETEN 220 (Polyquaternium-5), MERQUAT 550 (Polyquaternium-7) from Merck, and GAFQUAT 755 (Polyquaternium-11) from GAF. PREPARATION OF SAMPLES Virgin brown hair tresses, obtained from DeMeo Brothers, were given a simulated shampooing with TERGITOL © nonionic surfactant 15-S-9, thoroughly rinsed, and air dried. Polymer treatments were performed by immersing the tresses in an aqueous solution of the cationic polymer (0.1-0.01% by weight polymer) for 30 minutes with intermittent stirring. They were then washed with three 30-second distilled water rinses and air dried. In cases where a post-treatment with sodium dodecyl sulfate (SDS) was performed, the polymer-treated hair was immersed for 5 minutes in a 1 weight % SDS solution, again with intermittent stirring, and rinsed as above. ESCA ANALYSIS Hair fibers were analyzed using a Surface Science Laboratories Inc. SSX-100 ESCA spectrometer. Samples were mounted by suspending a V2-inch-long bundle containing 10-20 hair filaments across the 7 X 3-mm open area of a 10 X 13-mm Au-plated mask using double-sided adhesive tape. The masks were then suspended several milli- meters above a computer-controlled rotation platter. This effectively defocused the platter from the photoelectron acceptance optics, and therefore the void areas within the mask contributed no signal. The instrument is regularly calibrated to give the Au4f'7/2 photoelectron peak at 83.93 eV and the Cu2p3/2 peak at 932.47 eV. Since hair is nonconductive, all samples showed considerable charge buildup under x-ray irradiation. Analyses were therefore performed using an electron flood gun to provide charge com- pensation, and all measured binding energies were charge corrected to the principal C Is photoelectron line at 284.6 eV. Quantitation was performed using software and sensi- tivity factors supplied by the manufacturer. RESULTS AND DISCUSSION Because of its extreme surface sensitivity, ESCA would seem to be an excellent tool for studying the deposition of polymers on the hair substrate. A key consideration in this application, however, was the measurement reproducibility since this ultimately deter- mines the utility of quantitative analysis. Hair, being a natural product, was expected to be quite heterogeneous, and thus the measurement precision might also be expected to be rather poor, particularly given the small sample size used in these analyses. There- fore, a set of six samples of control hair (no polymer treatment) was initially analyzed in order to define the statistical variation inherent in the analysis. These control samples

ESCA OF POLYMERS ON HAIR 235 were obtained from the center portion of a single 8-inch-long tress. The results of these replicate analyses are given in Table I which shows the surface elemental compositions in atomic percent. As shown, the measurement precision is quite acceptable, with rela- tive standard deviations near 10%. Thus, heterogenity, at least at the spatial resolution of these analyses (2 mm2), does not appear to be a major obstacle. In addition to the survey spectra from which the elemental compositions were com- puted, high-resolution spectra of selected elements were also acquired. These spectra provide information on the types of functional groups, i.e., the chemical state of the elements present on the surface, since the binding energy of a photoelectron is per- turbed slightly depending on the nature of the bonding in which the particular nucleus is involved. Bonding which decreases the electron density around the nucleus produces a shift to higher binding energy. Thus, in the case of carbon, C- O, C- O, and CO0 (carboxyl) type groups shift the Cls photoelectrons to successively higher binding en- ergies. The contributions of each of these to the total observed peak envelope can be determined by peak-fitting analysis. The observed carbon peak envelopes of the control hair samples were well reproduced by fitting with three peaks (Figure 1). In order of increasing binding energy, they can be assigned to hydrocarbon, alcohol/ether groups, and amide carbonyls (2,5). Table II summarizes the results of this peak-fitting proce- dure as applied to the control samples. High-resolution nitrogen spectra revealed only one form of nitrogen to be present, with a binding energy consistent with that of an amide group. In agreement with the pro- teinaceous structure of the hair fiber, the concentration of the amide carbonyl carbon peak shows good agreement with the level of amide nitrogen detected. This is a good indication that the carbon peak-fitting parameters used were satisfactory. Two forms of sulfur were observed. These two forms have been observed in previous ESCA studies on hair surfaces and have been attributed to the presence of sulfonate and disulfide forms of sulfur (3). The replicate analyses of the untreated control hair serve as a benchmark by which to Table I Surface Composition of Control Hair Atomic % Sample C O N S Si A 78.6 11.8 6.2 2.3 1.0 B 76.0 13.4 6.2 2.6 1.0 C 78.1 11.5 7.1 2.4 0.9 D 74.7 13.9 6.7 2.5 1.1 E 74.8 14.3 5.3 2.1 2.4 F 78.0 11.8 6.8 2.9 0.3 X 76.7 12.8 6.4 2.5 S 1.8 1.2 0.6 0.3 R.S.D. 2.3% 9.4% 9.4% 12% X = average. S = standard deviation. R.S.D. = relative standard deviation.