344 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS one at the fiber tips. In contrast to this, Lunn and Evans reported the existence of only disentanglement peaks on tribocharge distribution curves (4). The static detector probe and the comb are located at different positions in the experimental set-up. Such an arrangement of sensors was necessary to minimize interference of the static detector probe from the electrical field of the charged comb. The data can thus be reported in graphs showing force and charge as a function of time of combing, as in Figure 2a, or as a function of distance along the tress length as in Figure 2b the latter allows for comparison of charge and force at the same point on the tress. Mechanical measurement of combing forces, represented by the lower curves in Figures 2a and 2b, reveals that the force is relatively constant for most of the tress length, rising significantly near the tip end. This is related to disentangling of "cross-over" hairs. High combing forces, arising from disentangling of tip ends, lead to better contact and increased friction between the comb and the fibers. These factors might contribute to the presence of a distinct tip-end peak in the charge distribution curves which corre- sponds to the maximum of the combing force curves shown in Figure 2b. The origin of the disentanglement peak in charge distribution curves was proposed earlier by Lunn and Evans (4). An explanation for the large charge peak in the upper portion of the tress may be related to the insulation characteristics of plastic combs and their ability to acquire a high electrical potential from charge accumulation. Combing of a hair tress causes contin- uous contact of an increasingly charged comb surface with fresh, uncharged portions of hair fibers. The conditions for electron transfer are most favorable in the immediate vicinity of the comb insertion point, where an uncharged or low potential comb con- tacts the keratin surface. As the process of electron transfer continues during the move- ment of the tress, the comb becomes charged to high potential and its surface states available for electron exchange are depleted. This should lead to an equilibrium in terms of electrochemical potentials of contacting surfaces and inhibition of electron migration (5,6). The charge density should, thus, be highest in the area close to the point where combing starts and drop to lower values in the middle section of a tress. The tribocharge density might rise again in the tip-end portion of a hair tress because of increased contact and friction associated with disentangling of "cross-over" hairs. This interpretation of experimental curves obtained with insulator combs is upheld by qualitatively different charge density distributions recorded in metal-comb electrifica- tion experiments depicted in Figure 2c the prominent peak in the upper portion of the tress is not present, and the charge distribution generally parallels the combing force curve. This result with the aluminum comb is consistent with the constancy of the comb and hair surface potentials throughout the whole combing test. Only when the entanglement at the tip end is reached does the increased friction lead to peaks in force and charge. This analysis of the charge distribution curves is, however, hard to reconcile with data from repeated combing with charged nylon comb. Figure 2d shows the distribution of charge after three consecutive combings which involved a discharged tress and a charged comb. After the first combing, prior to which both the comb and the hair were dis- charged, the intensity of the peak in the upper portion of the tress was relatively low. The second and the third combing resulted in an increase in the intensity of the first peak. The charge densities corresponding to the lower portions of the hair tress were

TRIBOELECTRIC CHARGE DISTRIBUTIONS ON HAIR 345 similar after each combing, independent of the comb potential. This effect of enhanced charging by the probe loaded to high potential of the opposite sign is difficult to explain within the framework of existing models of polymer-polymer or metal-polymer electrification. According to the band model of contact charging proposed by Davis (5,7) and Lewis (8), the charge would be transferred until either all the surface states are filled, or until a sufficient surface potential is created to prevent further charge transfer. Repeated contact usually leads to accumulation of transferred charges, which is explained by slow diffusion from surface states into bulk states, creating surface vacancies which can be subsequently refilled. This additional charge transfer is, thus, related to the concentra- tion gradient of charged species within the bulk and to the rate of transport of the surface states to the bulk states. For insulators such as comb materials used in this study, and keratin with low moisture content, the dielectric relaxation times are long, and consequently rapid filling of the surface states and slow filling of bulk states can be expected. This representation of contact charging justifies the enhanced electron transfer in the upper part of a tress but fails to explain why a charged comb contributes to the further intensification of this process. Another theory of triboelectric charging, the Duke and Fabish (9) sampling/non-communicating state model of polymer-polymer contact, does not account for the surface potentials developed after charge injection and does not predict how the state energies might be modified by the existence of electrical fields created by the excess charge. It cannot be thus used to analyze the effects reported in this paper. The middle peak in the charge distribution profiles shown in Figures 2a and 2b is sometimes undetectable in charge distributions, or it becomes merged with the entan- glement peak. Some data suggest that it might be an artifact which we believe may be related to the change in geometry of the tress as the hair clears the comb. However, the measurements of charge distribution without combing on previously charged tresses also demonstrate the existence of this peak. EFFECT OF COMB WORK FUNCTION AND MULTIPLE COMBINGS Teflon (4.26- 6.71 eV), polyethylene (4.9- 6.04 eV), nylon (4.08- 4.5 eV), polycar- bonate (3.85-4.8 eV), and aluminum (3.38-4.25 eV) combs were used to assess the effect of comb material on charge distribution profiles. The range of work function values (the work required to remove electrons from the Fermi level to the surface) given in brackets and reported in the literature serves only as a general indication of relative positions of these materials in the triboelectric series (3). As mentioned in our previous paper, there is a considerable discrepancy between various sources, mainly due to the use of different experimental procedures and materials with varying degrees of purity. Figures 3a-e shows charge density distributions on hair tresses combed with nylon, polyethylene, teflon, polycarbonate, and aluminum combs. The same figures illustrate the gradual buildup of static charge on hair as a result of consecutive combings. The numbers assigned to each distribution curve represent'the charge densities integrated over the length of the tresses (expressed in C/cm). Figure' 3a presents the charge distri- butions obtained with a nylon comb, including the combing force curve corresponding to the first combing cycle. The shape of the combing force curve as well as the values of the forces are representative for all comb materials studied. This is in accord with the