SKIN PENETRATION 627 accelerant causes the keratin to swell and leaches out essential structural material from the stratum comeurn, thus reducing the diffusional re- sistance and increasing the permeability (31, 58, 91-93). Varied agents have been reported to have accelerant action, particularly propylene glycol, surface active agents, and aprotic materials such as urea, DMSO, DMF, and DMA. The evidence for the effect of propylene glycol in skin penetration is conflicting (56, 62, 69, 77, 88, 94-100). In combination with surface active agents, propylene glycol was claimed by MacKee and coworkers (94) to promote the penetration of water-soluble substances via the transfollicular route. A concentration of 1% 5-fluorouracil in propylene glycol has been found to be as effective clinically in certain skin car- cinomas, as opposed to a minimum concentration of 5% in an ointment (100). Enhanced effects were noted in absorption of varied steroids (56, 62, 77, 96, 98, 99). The most effective of the "accelerants" are DMSO, DMF, and DMA. On application to the skin, DMSO passes rapidly through the stratum corneum (52, 101-104) and can aid in the penetration of a wide range of substances. The role of DMSO in enhancing percutaneous migration is well documented (13, 91, 101, 105-110). DMF and DMA enhance cutaneous penetration to a lesser degree than DMSO. Work with these solvents is contributing greater understanding of the chemical nature of the skin barrier in relation to a specified penetrant and the transport mechanisms of various compounds across skin. The mechanism of action of these penerrant accelerants needs greater definition. DMSO, DMF, and DMA are all strongly hygroscopic and it is likely that the presence of these substances in the stratum comeurn in- creases the hydration of the tissue and therefore its permeability. There are possible reversible configuration changes in skin protein structure brought about by substitution of integral water molecules by DMSO (111), with resultant swelling. This may explain, in part, the sometimes observed variable role of DMSO. Swelling may induce the formation of channels within the matrix of the stratum comeurn which either favor the passage of varied compounds (59) or lower the diffusional resistance of the stratum comeurn. DMSO can extract soluble components from the stratum comeurn, suggesting ultrastructural modifications consistent with an increase in permeability. DMSO exhibits an unusual concentra- tion dependence (15, 52, 58, 59, 112). At least 60% DMSO is required for a measurable penetration rate. The rate rises to a maximum at

628 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS about $ hours and then declines, and this is true for both the pure liquid and for DMSO-water mixtures containing 70% DMSO. The con- tinued increase in the penetration rate of pure DMSO with increase in the applied quantity is probably a consequence of progressive barrier impairment. The decline in the peak rate after 3 hours is probably due to the back-diffusion of water with consequent dilution of the DMSO (52). Surface active agents appear to play a major role in promoting trans- appendageal absorption. The influence, particularly of anionic type ma- terials, seems related to their ability to increase the permeability of the skin to water by altering the physical state of water in the skin in such a way as to permit greater freedom to the passage of charged hydrophylic substances (73). The irritant action of anionic materials, such as, e.g., "soap" or sodium lauryl sulfate, suggests that they must penetrate to susceptible tissue. The role of surfactants in percutaneous absorption has been reviewed by Barr (8d), Sprott (75), Ritschel (107), Scala et al. (113), and Minato et al. (63). Penetration of certain antimicrobial substances appears enhanced by addition of surface active agents. Washing with sodium dodecyl sulfate enhanced the amount of hexachlorophene and tetrachlorosalicylanilide which penetrated rat skin (73). However, the bulk of evidence indicates that the stratum corneum is an effective barrier (12, 14, 59, 112-116). When penetration occurs, anionics penetrate best (12, 14, 113, 115-117), followed by cationics and nonionic surfactants (118, 119). Among anionic substances the laurate ion is reported to have the greater penetration and the most effect on the penetration of other solutes (117, 120). Soaps of different fatty acids have this property in varying degrees (117, 119), with penetration more significant for salts of fatty acids with chain length of 10 carbon chains or less (14, 73, 115). The penetration of fatty acid soaps varies inversely with pH (14, 73, 115). At higher pH (ca. 11), the action of the anionic surfactant appears to be attenuated or overshadowed by the influence of the more alkaline pH itself. Interpretation of surfac- rant action upon the skin must deal separately with these two phenomena, i.e., pH on the one hand and surfactant effect on the other. Possible mechanisms whereby the skin is able to restrict the percutaneous migra- tion of synthetic anionic surfactants have been reviewed by Blank and gould (14). Scala and his colleagues (113) calculated permeability constants for a wide variety of materials, including surfactants. The nonlinearity of diffusion curves (permeability constants vs. time) for an anionic surfac-