70 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Figure 14. TEM micrograph showing a rather thin deposit of prismatic TiC) 2 crystals along the outer horny layer of human skin. The scratches (arrows) running perpendicular to the corneocytes (Co) correspond to sectioning artefacts. Sunblock preparation 3 (scale bar = 100 nm). prismatic (Figure 3) habits. The patterns of minute parallel striations within some TiO 2 particles reflect ordered structures, thus demonstrating the crystalline nature of the employed sunscreen agent (Figure 3). TEM micrographs of the acicular form of TiO 2 (Figure 2), prior to formulation, showed a tendency of the tiny crystallites to gather in groups of four or five individual crystals, resulting in the formation of larger particles. Such arrangements may have the drawback of converting ultrafine titanium dioxide into pigmentary TiO 2 particles. X-ray diffraction analysis allowed us to verify the stoichi- ometry and crystallographic structure of the selected raw materials. Indeed, all peaks in the X-ray spectrum could be indexed, corresponding either to the rutile form or to the anatase form of titanium dioxide (Figure 4). The technique employed for collecting the outer layer of the stratum corneum is similar to the adhesive-coated discs used in previous studies (19). The plastic substratum we used gave satisfactory results in ultramicrotomy. The spherical aspect of the cosmetic vehicle (H•liosides ©) used in sunscreen preparations 1 and 2 could be clearly identified in SEM and TEM (Figures 5,6,8). However, the inner organization of the mineral agent within the microspheres differed markedly. In fact, the acicular form of TiO 2 (sunscreen preparation 1) seemed preferentially confined at the boundaries of the oil droplets (Figure 6), whereas the prismatic titanium dioxide (sun- screen preparation 2) literally filled the spherical structures (Figure 8). A possible explanation of these different arrangements may lie in the surface treatments. These latter treatments may determine whether the ultrafine oxides are hydrophilic or hydro- phobic. It seems noteworthy to mention that despite the important specific surface area

PHYSICAL SUNSCREENS 71 (up to 50m2/g -•) of the ultrafine TiO2, the crystallites remained nonagglomerated and loosely packed in the microspheres (Figure 8). Although the various titanium dioxide raw materials (acicular or prismatic) presented different spatial arrangements within the oil droplets, the TiO 2 crystals appeared to be mainly encapsulated in those round bodies (H•liosides©), thus isolated from the aqueous matrix. After topical application to skin, the oil droplets, containing the physical sunscreen, may burst and produce a quite homogeneous distribution of the TiO 2 crystallites along the stratum corneum (Figures 7,9-11). No particular microstructure was evident in sunblock preparation 3. The titanium dioxide crystals occupied the whole cream layer (Figure 12). The mineral ingredient seemed nonagglomerated and widespread. After application of sunblock preparation 3 onto the skin, either thick (Figure 13) or thin (Figure 14) crystalline coatings of the outer horny layer could be noticed. In spite of the different crystal arrangements (Figures 6, 8) within the oil droplets (H61iosides©), both sunscreen preparations 1 and 2 present similar spatial distributions of the TiO 2 after application to skin, that is a thin and almost monocrystalline coating of the stratum corneum (Figures 7, 10). These thin oxide layers seem suitable because of the organic filter/mineral screen synergy offered by those formulations. On the other hand, sunscreen preparation 3, prepared solely with physical sunscreen agents without any chemical sunscreen added, generally showed the inorganic powder randomly dis- tributed within a thicker layer (Figures 12, 13) compared to sunscreen preparations 1 and 2. Furthermore, it must be pointed out that the ultrafine titanium dioxide crys- tallites appeared to be exclusively confined to the outer surface of the stratum corneum and around desquamating corneocytes (Figures 10, 13). In summary, our observations provide conclusive evidence that the spatial distribution of a physical sunscreen is greatly influenced by the formulation of the sun care product. Thus, the sunscreening efficacy is directly related to both the mineral ingredient and the formulation. Moreover, this study emphasizes the contribution of electron microscopy in determining the location and behavior of TiO 2 crystals upon the skin's surface. Such electron-microscopic assessments yield useful information for quality checks of physical sunscreens, thus allowing us to explore product effectiveness. REFERENCES (1) S.-Y. Lin and R.-C. Liang, The effect of ultraviolet B irradiation on the isolated porcine stratum corneum: Colorimetric and ATR/FT-IR spectroscopic investigations, Biotaed. Res., 15 (1) 9-15 (1994). (2) L. A. Applegate and E. Frenk, Cellular defence mechanisms of the skin against oxidant stress and in particular UVA radiation, Eur. J. Dermatol., 5, 97-103 (1995). (3) M. F. Naylor, A L Boyd, D. W. Smith, G. S. Cameron, D. Hubbard, and K. H. Neldner, High sun protection factor sunscreens in the suppression of actinic neoplasia, Arch. Dermatol., 131, 170-175 (1995). (4) R. Marks, P. A. Foley, D. Jolley, K. R. Knight, J. Harrison, and S. C. Thompson, The effect of regular sunscreen use on vitamin D levels in an Australian population, Arch. Dermatol., 131, 415-421 (1995). (5) R. M. Sayre, N. Kollias, R. L. Roberts, A. Baqer, and I. Sadiq, Physical sunscreens, J. Soc. Cosmet. Chem., 41, 103-109 (1990).