OW MICROEMULSION IN SUNSCREENS 457 be obtained. This ternary system gave the most efficient microemulsion, with a 9.0% disperse phase and a 14.67% surfactant mixture. Among the systems prepared with cetearyl sulfate and soya lecithin, the best was probably that with C•2_•5 alkylbenzoate/soya lecithin/decylpolyglucose, water, and 1,2 hexanediol, since it had the lowest percentage of surfactants and a good amount of disperse phase, which could be partially replaced with a sunscreen agent to give solar products a suitable protective effect. Since the primary aim of this study was to obtain transparent, nonsticky, and waterproof sunscreen products, sprayable and with a fresh feeling on the skin, in a second step of the study, the lipid phase of these systems was partially replaced with 4-methylbenzilidene camphor, a crystalline powder, and octylmethoxycinnamate, an apolar oil, both of which are sunscreen agents frequently used in cosmetic products. For this purpose, it was necessary to use a lipid phase that could easily solubilize both sunscreens, to avoid areas of crystals or of oil separation, which can originate from non-uniform products with poor diffusion on the skin, or areas of excessive concentration of sunscreen. To avoid these unwanted effects, ethanol was mixed to the lipid phase. Two ternary diagrams were then constructed for each sunscreen agent, in which a 2/1 or 4/1 mixture of C•2_•5 alkylbenzoate/ethanol was at one corner, cyclomethicone was at the second corner, and octylmethoxycinnamate or 4-methylbenzilidene camphor was at the third. Cyclomethicone was used to provide a fresh feeling and nonsticky products, and ethanol was used to increase the compatibility of C•2_•5 alkylbenzoate and cyclo- methicone and sunscreen miscibility in the systems. Figures 5 and 6 give some examples of ternary diagrams of the lipid systems constructed, respectively, with octylmethoxycinnamate, C•2_• 5 alkylbenzoate/ethanol (4/1), and cy- clomethicone (Figure 5) and with 4-methylbenzylidene camphor, C•_• 5 alkylbenzoate/ ethanol (2/1), and cyclomethicone (Figure 6). The absence of crystals or phase separation was checked with a light polarized microscope. C12-C16 alkylbenzoate/ethanol 4tl Octy Im etho xy c innam ate Figure 5. Ternary diagram of systems with C•2_•5 alkylbenzoate, ethanol, cyclomethicone, and octylme- thoxycinnamate.

458 JOURNAL OF COSMETIC SCIENCE alkylbenzoatelethanol 2/1 4-Methylbenzilidencanpho r •l• Cyclomethicone Figure 6. Ternary diagram of systems with C12_15 alkylbenzoate, ethanol, cyclomethicone, and 4-meth- ylbenzilidene camphor. The best systems were those obtained by mixing 90%, 80%, and 70% of the 10/90, 20/80, 30/70, or 40/60 cyclomethicone/(C•2-C•5 alkylbenzoate/ethanol 2:1) mixture with methylbenzilidene camphor. Octylrnethoxycinnarnate had a higher affinity for the lipids than did 4-methylbenzilidene camphor and required less ethanol for its dissolu- tion. Two transparent systems with high percentages of sunscreen agents were chosen to avoid their excessive dilution in the final microemulsions: A: 32.7% C•2_•5 alkylbenzoate, 16.3% ethanol, 21.0% cyclomethicone, 30.0% 4-meth- ylbenzilidene camphor. B: 33.6% C•2_•5 alkylbenzoate, 8.4%, 18.0% cyclomethicone, 40.0% octylmethoxy- cinnamate. These solutions, with the highest percentages of lipids and sunscreens, were chosen to reformulate microemulsions with good amounts of water, and pseudoternary systems were drawn to prepare more O/W microemulsions with water:l,2-hexanediol (2/1) or water:2-methyl,2,4-pentanediol (4/1) as polar phase and decylpolyglucose/soya lecithin as surfactant. As an example, a ternary diagram of the mixtures with 2-rneth- ylbenzilidene camphor and with 1,2-hexanediol as cosurfactant is presented in Figure 7. Table III shows the compositions of transparent systems with 4-methylbenzylidene camphor. Figure 8 shows the ternary diagrams of the system with octylmethoxycinnamate in the presence of water/1,2-hexanediol 2/1. The compositions of microemulsions with octyl- methoxycinnamate are shown in Table IV. By using octylmethoxycinnamate with 1,2-hexanediol and water as polar phase, the replacement of the lipid phase with the complex system changed the existence area and gave systems with the lowest percentage of the surfactant/lipid mixture. With 2-rnethyl- 2,4-pentanediol as polar phase, the mixture of 60% lipid/surfactant to 40% polar phase was not obtained. The microemulsion with octylmethoxycinnamate chosen was the system obtained by diluting 50% of the 30/70 lipid/surfactant mixture with 50% polar phase, with hexanediol as cosurfactant. The microemulsions with 4-methylbenzylidene camphor at 4.5% and octylmethoxy- cinnamate at 6.0% had the following compositions: