0 - C C - STABILIZATION OF 1-ASCORBIC ACID 1.0 -•--•--•-• -----t 0.8 -■- 1 wt.% L-ascorbic acid -•- 3 wt.% L-ascorbic acid -•- 5 wt.% L-ascorbic acid 0.6 0.4 0.2 0.0-t--....--.......-....---,--....... -....... ----------.... 0 2 4 6 8 Storage time (week) 10 9 Figure 5. Stability of L-ascorbic acid in w/o/w double emulsions with the storage time (40°C) at different concentrations: - ■-, 1 wt% L-ascorbic acid -•-, 3 wt% L-ascorbic acid -A-, 5 wt% L-ascorbic acid. All the samples were shielded with 0. 5 wt% MgSO4. The pH of the internal aqueous phase was adjusted to 7. is established, 1-ascorbic acid can be stabilized to a high concentration in the w/o/w double emulsions. A more attractive result in our study was that the 1-ascorbic acid stabilized showed no browning phenomenon. Figure 6 shows a photograph for the color change of 1-ascorbic acid-loading w/o/w double emulsions. Due to the deformation of chemical structure, the color of emulsion formulations changes inevitably from a white to a dark brown, which is accelerated at elevated temperatures. Such a phenomenon is a fatal defect in the final applications. In order to confirm the browning effect of w/o/w double emulsions, the (a) (b) Figure 6. Color change for L-ascorbic acid-loading w/o/w double emulsions after 10-week storage at 40°C: (a) pH 2, (b) pH 7. The concentration of L-ascorbic acid in the emulsions was fixed at 3 wt%.

10 JOURNAL OF COSMETIC SCIENCE color change was observed at an elevated temperature, 40 ° C. In the case of a low pH, the color of w/o/w double emulsions changed eventually to dark brown (Figure 6a). The browning started to appear within several days. However, when the pH of the internal aqueous phase was adjusted to neutral, we could not observe any change in color (Figure 66). This result is also direct evidence for the successful stabilization of 1-ascorbic acid by means of ionic shielding in the w/o/w emulsion system. SKIN PERMEATION CHARACTERISTICS OF L-ASCORBIC ACID Skin permeation of 1-ascorbic acid is very important for dermatological applications. In order to show its functions properly, 1-ascorbic acid must permeate the stratum corneum and reach viable epidermal and dermal layers or viable subcutaneous layers. However, the barrier property of skin acts as a major obstacle to cransdermal drug delivery. Figure 7 shows the in vitro skin permeation of 1-ascorbic acids. It was observed that the skin permeation efficacy of 1-ascorbic acid in the neutral condition is significantly lowered compared with that of the acidic condition. This is because at the neutral condition, the skin, especially stratum corneum, is not injured, as compared to the acidic condition. Also, the lower skin permeation at the neutral condition stemmed possibly from higher ionization, that is, the higher polarity of 1-ascorbate ion at the neutral condition. Nevertheless, even though the efficacy was more or less low, it is significant that the skin permeation of 1-ascorbic acid still took place successfully. Ionic shielding seems to have negligible influence on skin permeation. CONCLUSIONS In chis study, we tried to stabilize 1-ascorbic acid in w/o/w double emulsions. From the fundamental understanding of the behavior of 1-ascorbic acid in the aqueous solution, - N E - -, 3000 2500 2000 1500 1000 500 0 AA (pH2) AA (pH7) A�gS0 4 (pH7) Figure 7. Evaluation of 1-ascorbic acid in vitro skin permeation with Franz diffusion cell measurements.