NEW EVALUATION TECHNIQUES FOR SUNSCREENS 589 most efficiently, and the greatest divergence from the values occurred where the physical particles exerted the greatest effect in the total ab- sorption of light (i.e., near 360-700 m• and 230--270 m•). The com- parable runs without chemical absorbers and also without the physical sunscreen show even greater divergency between measurements, as shown in Fig. 3. Again this shows the loss of light due to scattering by physical particles rather than to normal absorption. The photocell- monochromator combination just could not detect as much of the scattered light as the integrating spheres. Some difference in the shape of curves obtained by the different tech- niques is to be anticipated from the optical and mechanical inferiority of the monochromator as compared to the Cary spectrophotometer. Two important reasons are: (a) the intensity of the light source of the monochromator is unstable it is not simultaneously in balance with a reference beam (b) the monochromator optics are much poorer (wider slits, greater band width, etc.) since it was not designed for narrow-band spectrophotometric work. Using films of different thicknesses, 0.0254, 0.0019, 0.0127, and 0.00635 mm + 10%, with physical and chemical/physical sunscreen products, it was found that for transmission measurements the physical sunscreens do not obey the Lambert-Beer Law. The reflective characteristics of the thin films were also examined to determine the degree to which reflection contributes to the protective ability of sunscreen products. In Fig. 8 the reflection spectra of films with physical and chemical/physical sunscreen products may be seen. Figure 2 shows the reflection spectrum of a film with a chemical sun- screen. The most striking feature of the reflection spectrum is that it appears to follow the total transmission spectrum. If the sample trans- mits well it also reflects well. If it absorbs well in some spectral region, or transmits little light, it does not reflect well. For example, the opaque cream base reflects uniformly at all wavelengths slightly more light than the p-aminobenzoic acid (clear viscous) solution. However, addition of 10% talc to the opaque cream base doubles its reflective capacity in the 300 m/• to 700 m/• region. Addition of 5% amyl p-di- methylaminobenzoate to the talc product did not increase the reflection very much. On the contrary, because of the absorbing capacity of the amyl-p-dimethylaminobenzoate in the sunburn range, the reflection of the talc product was reduced to the level of the opaque cream base in the 230 m• to 350 m/• range from 350 m/• to 700 m• the reflection was equal to or slightly higher than that from the talc product.

590 JOURNAL OF THE SOCIETY OF COSMETIC C[fEM1STS Comparison of monochromator or integrating spheres measurements of thin films with the measurements of the solution-dilution method re- veal interesting and important observations: The transmission char- acteristics are qualitatively similar for the chemical sunscreen product as determined by both methods (Figs. 1 and 2) however, these are entirely different for the physical and the chemical/physical sunscreen products, particularly in the longer wavelengths (Figs. 2 and 3). The physical sunscreen, being insoluble in solvents used in the solution-dilution meth- od, obviously cannot affect the transmission spectrum as it does in the IOO 9o 8o 7o 6o 50 4o 30 20 IO WAVELENGTH ml• Figure 8. Percent reflection using spectrophotometer with integrating spheres of 0.0254 tnm thick films of three sunscreen products. -- 5% amyl-p-dimethylaminobenzoate q- mrs/ talc in opaque cream base - - - 10% talc in opaque cream base and . .. opaque cream base film specimens. This also is responsible for the lack of reflection in the solution-dilution method samples, although reflection is observed so well in the thin film specimens. Naturally, the lack of physical sunscreens in the extracted solution makes it impossible to determine the differences in transmission when one physical sunscreen is added to another. In conclusion, measurements with the thin film technique are more representative of conditions actually occurring on the skin since the product, when applied on the skin, will form a film, and the thickness of this film can be reproduced by this technique. In addition, the product is optically measured quickly and conveniently in its original form with- out being laboriously altered to the solution form, as in the solution-dilu- tion method.