PHOTOACOUSTIC MEASUREMENTS ON SKIN 381 _1,0 •O.U. •m '05 • ß I I Euso[ex 6300 in Isopropano[ Concentrotion 1.0ø/o 0.5ø/ø Eusotex 8020 1.0 ø/o 0 I 250 300 f=1200 Hz 350 nm 400 Wovetength Figure 4. Photoacoustic i. vivo spectra of Eusolex 6300 and 8200 topically applied to the skin in isopro- panol solution. The curves show the differences in signal amplitudes between treated and untreated skin. tissue from which it may be absorbed into the blood capillaries. The protective action against sunburn continues to exist as long as the agent remains on the skin surface or within the stratum corneum (12). Apart from external removal of the preparation, two processes are of special interest for the temporal variations of the distribution of sun- screen on and in the horny layer: the transport of the agent to the horny layer, de- pending on the vehicle/stratum corneum partition coefficient, and the diffusion through the horny layer, characterized by the diffusion coefficient of the sunscreen in the stratum corneum. In many cases the latter process comes out to be the rate-determining process for the penetration of the agent into the skin. As an example, Figure 5 shows experimental results obtained in vivo with Eusolex 8020 at 350 nm, close to the wavelength where this UVA filter shows maximum absorption. In comparison to the solution used for the determination of the differential spectrum of Figure 4, in this experiment an isopropanol solution of one tenth the concentration was

382 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 1.0 I I I ß I ß f = 180 Hz ß ""' ""' i - -• o • I - r r t Eusotex 8020 ),=350 nm h Time Figure 5. Time dependence of the photoacoustic in vivo signal of the skin after topical application of Eusolex 8020 in isopropanol solution (concentration 0.1%). applied to the skin. The plotted curves, measured at chopping frequencies of 180 and 1,200 Hz, show the photoacoustic signal before and after the application of Eusolex 8020 and its dependence on time. At the chosen chopping frequencies the open-ended differential detector employed has the same sensitivity. The larger experimental error of the 180 Hz data (_-4- 15% against _ 4% at 1,200 Hz) is caused by the frequency-depen- dent acoustic impedance of the skin surface terminating the open-ended detector. At low chopping frequencies the sensitivity of the detector shows a stronger dependence on the force with which the forearm is pressed to the cell rim (13). The values of the thermal diffusion length are 12 and 4 Ibm at the respective chopping frequencies. Therefore, the absorption of light within the whole horny layer or within only the superficial corneal layers, respectively, contributes to the photoacoustic signal. The increase in signal amplitude observed immediately after application of Eusolex 8020 is the same at both chopping frequencies. At the higher frequency the signal rapidly decreases to the value of untreated skin with a half-life of about 3 h. This means that after this time approximately 50% of the agent has penetrated into the horny layer to a depth greater than the thermal diffusion length of 4 tzm. The weak decrease of the 180 Hz signal, on the other hand, shows that within the observation period of the experi- ment most of the sunscreen is still inside the horny layer. A detailed discussion of the frequency dependence of the photoacoustic signal in terms of the concentration profile