PHOTOACOUSTIC MEASUREMENTS ON SKIN 383 of a light-absorbing agent's penetration into the horny layer has been given elsewhere (ii). The investigation of the penetration of Eusolex 8020 into the stratum corneum for which Figure 5 gives a representative example was performed with four test subjects. The results show low interindividual scatter. For the half-life of the signal decrease at 1,200 Hz, a mean value of 2.9 - 0.9 h was obtained. F •m the theory of diffusion it is found that the amount of agent within the superficial layer of thickness e, which is detected by the photoacoustic signal, decreases in proportion to the error function of argument e/2X/Dt as long as most of the agent is still inside the horny layer. Here t denotes the time after application of the sunscreen. Correspondingly, the half-life of the photoacoustic signal at 1,200 Hz yields a mean diffusion coefficient D of Eusolex 8020 in the stratum corneum of the forearm of about 1.7 x 10 -• cm2/s. This value is smaller by a factor of about 6 than that reported by Scheuplein for the diffusion of low molecular weight molecules in the dry stratum corneum of the forearm (14). Different results were obtained with Ilrido © cream which in this study was the prepara- tion with the largest concentration of sunscreening agent. The initial strong photo- acoustic signal increase after cream application, with a further weak increase during a postapplication period of about one hour, was in this case followed by a much slower signal decrease. In Figure 6 the mean photoacoustic signal as obtained by averaging the 1.0 Oo go 0.5 I I I I Appticotion of I[rido Creorn 300 n rn 1200 Hz 0 2 /, 6 h 8 -,- Time Figure 6. Time dependence of the photoacoustic in vivo signal of the skin after topical application of Ilrido © cream.

384 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS individual 1,200 Hz signals of six test subjects is plotted over time. A measurement performed after a postapplication period of 24 h showed the sunscreen signal to be plainly visible. The half-life was estimated to be about 15 h. The 180 Hz signal showed a very similar temporal behavior, so that with the Ilrido © cream the transport of the sunscreen from the vehicle to the horny layer seems to be the rate-determining process for the penetration of the agent into the skin. For the weak signal increase observed during the first hour after application, two opposing processes might be responsible: (i) the penetration of the sunscreen into the stratum corneum and (ii) the redistribution of the preparation from the sulci over the skin surface. While the first process is respon- sible for the decrease of the photoacoustic signal, the latter yields an increase as the amount of agent which is reached by the light and contributes to its absorption is increased. Recently Blank eta/. observed in vitro a similar increase of the absorption ability of a sunscreen during the first two hours after topical application to excised stratum corneum and interpreted this effect as being caused by a redistribution of the agent (12). The authors were able to demonstrate in vivo the corresponding delay in erythemal response of the skin to ultraviolet irradiation. CONCLUSION This study shows that photoacoustic spectroscopy can be used to observe in vivo the penetration of topically applied drugs into the skin if their ability to absorb radiant energy is sufficiently strong with respect to that of stratum corneum. At chopping frequencies in the audio frequency range, the method mainly yields information about the absorption of light on and within the horny layer. To obtain data about the viable epidermis would require much lower chopping frequencies. The stability of the agent against irradiation is of utmost importance as the photoacoustic in vivo measurements are performed at relatively high light intensities in order to achieve a satisfactory signal-to-noise ratio. Experimental results are presented which show the different pene- tration behavior of two sunscreening agents incorporated in different vehicles. As ex- pected, the penetration of the agent dissolved in alcohol occurred faster than that of the agent incorporated in a cream. ACKNOWLEDGEMENTS Financial support by the Deutsche Forschungsgemeinschaft, Bonn-Bad Godesberg, FRG, and the Paul G. Unna-Stiftung, Diisseldorf, FRG, is gratefully acknowledged. REFERENCES (1) R. C. Wester and H. I. Maibach, "In Vivo Methods for Percutaneous Absorption Measurements," in Percutaneous Absorption, R. L. Bronaugh and H. I. Maibach, Eds. (Marcel Dekker, New York, 1985), pp 245-249. (2) A. Rosencwaig, Photoacoustic Spectroscopy (J. Wiley, New York, 1980). (3) T. A. Moore, "Photoacoustic Spectroscopy and Related Techniques Applied to Biological Materials," in Photothem. Photobid. Rev. 7, K. C. Smith, Ed. (Plenum Press, New York, 1983), pp 187-222. (4) A. Rosencwaig, Potential clinical applications of photoacoustics, C/in. Chem., 28, 1878-1881 (1982). (5) I. Simon, A. G. Emslie, C. M. Apt, I. H. Blank, and R. R. Anderson, "Determination of In Vivo