140 JOURNAL OF COSMETIC SCIENCE Italian law has recently prohibited the use of hydroquinone in cosmetics (4), owing to its skin-irritant effect and to its potential involvement in carcinogenesis, so that most formulators are now interested in the use of naturally occurring whitening agents such as kojic acid and arbutin. Kojic acid, produced by many species of Aspergillus and Penicillum in an aerobic process (5 ), possesses an excellent whitening effect, owing to tyrosinase inhibition activity that is probably due to a copper chelating action. It can also prevent serious sunburn caused by an accumulation of melanin in subcutaneous tissue produced via a tyrosinase­ catal yzed metabolic pathway (6). Arbutin is an active ingredient found in plants such as bearberry (Arctostaphylos uva-ursi), and has long been used in Japan as a chemotherapeutic agent against a number of pigmentary disorders (7). Although its action mechanism is probably similar to that of hydroquinone, it is much less cytotoxic to melanocytes in culture (8). Unfortunately, both arbutin and kojic acid may undergo photo-and/or thermodegradation in aqueous solution (9), involving modifications in their chemical structure, loss of organoleptic properties, and darkening of formulations, making it difficult to obtain stable and effective products. In this experimental study, with the aim of enhancing the UVB photostability of both whitening agents, cosmetic microemulsions obtained with mild, naturally derived sur­ factants (10) were formulated and characterized. Some perfume formulations, developed and characterized in our research laboratory (11) using fragrant molecules belonging to different chemical groups, were also introduced in the microemulsions to evaluate the olfactory impact of the perfumed microemulsion and the effect of the fragrances on its stability in the presence and absence of whitening agents. The influence of perfumed compositions on the photostability of arbutin and kojic acid in microemulsions was also investigated. EXPERIMENT AL METHODS MATERIALS Microemulsions. Arbutin, kojic acid, 1,2-hexanediol, isopropylpalmitate [IPP], and 2-methyl-2,4 pentanediol (hexylene glycol) were from Aldrich sodium hydroxide and methanol were from Carlo Erba tetra hexyl ammonium chloride was from Sigma absolute ethanol and hydrochloric acid were from Fluka soya phosphatidylcholine 96% (lecithin) {SPC] Epikuron®200 was from Lucas-Meyer decyl polyglucose Oramix®NSIO (decyl polyglucose), a.s. 0.55%, was a gift from Seppic. Fragrances • Coconut: Ethanol 68.5% water 29.6% -y-octalactone 1 % -y-butyrolactone 0.5% vanillin 0.4%. • Fragrance: Ethanol 55.8% hydroxycitronellal 14.26% tetrahydro linalool 14.26% oak musk abs. 9.0% tibetolide 3.56% Turkish rose e.o. 1.8% a-hexyl cinnamic aldehyde 1.24%. • Orange: Orange 1/10 e.o. 45.0% sweet orange e.o. 45.0% orange terpenes 4.4% ethyl butyrate 1.5% aldehyde CS 1.04% orange heart 0.731 % isoamyl butyrate 0.55% lauryl alcohol 0.475% a-ionone 0.379% acetaldehyde 0.27% linalool

PHOTOSTABILITY OF WHITENING AGENTS 141 0.22% citral 0.12% �-ionone 0.066% mandarin heart cone. 0.052% ethyl caproate 0.029% tagette e.o. 0.0219% citronella! 0.015% ethyl pelargonate 0.015%. • Tea: Ethanol 93.8% green tea extract on supercritical CO2 5.4% cloves e.o. 0.2% fenugreek abs. 0.2%, ethyl caprylate 0.00364%, �-damascone 0.00364%, linalyl acetate 0.002% acetaldehyde 0.00182% terpinyl acetate 0.00182% ethyl caproate 0.00182% geranyl acetate 0.00182% isovaleric aldehyde 0.00182% neryl acetate 0.0018% linalool oxide 0.00164% linalool 0.0012% 4-ethylguaiacol 0.00091 % cis-jasmone 0.00076% trans-2-hexenal 0.0003% sandalwood e.o. 0.00028% meth­ yl salicylate 0.00024% terpinene-4-ol 0.00024%. • Violet: Ethanol 94.4% methyl a-ionone 2.0% violet leaves abs. 10% sol. 1.6% a-ionone 1.0% linalool 1.0%. All percentages are w/w. Apparatus. Devices employed were an HPLC apparatus cons1strng of a SPD-10 AV UV/vis. detector, an LC 9A pump unit control, and a C-R6A Chromatopac integrator (Shimadzu) sub-micron particle analyzer model N4 MD (Coulter Electronics) anchor stirrer DLS (Velp) UV/vis. Lambda 2 spectrophotometer (Perkin Elmer) and UVB lamp TL40W/12RST40Tl2 (Philips). MICRO EMULSIONS Preparation. Several alkyl glucoside surfactants were tested in microemulsion formula­ tions. They were previously dispersed in water then the chosen oil was added, tritating to transparency with the cosurfactant. In two separate series of experiments, kojic acid and arbutin were added to the chosen microemulsion at 0.25% w/w. Citral and linalool were also introduced in the same microemulsion to evaluate the influence of model odorous molecules on the photosta­ bility of both whitening agents. Characterization. Microemulsions in the absence and presence of whitening agents were tested for stability by applying repeated freeze-thaw cycles (4 hours at -20°C, 4 hours at +40°C, and 16 hours at room temperature for 1 week). Mean diameters were then determined at 25 .0° ± 0.1 ° C by means of quasi-elastic laser light scattering techniques (QELLS), as described elsewhere (12). PERFUMED MICROEMULSIONS Preparation. The fragrances coconut, fragrance, orange, tea, and violet were added to the previously prepared and characterized microemulsion in the 0.01-0.1 % w/w range, after diluting them with ethanol. In microemulsions containing 0.03% w/w of each fragrance, 0.25% w/w kojic acid or arbutin was added in separate experiments. Characterization. All measurements were performed as described for microemulsions in the absence of fragrances. Olfactory evalt1ation. Olfactory analysis on perfumed microemulsions was done as follows: 20 ml of the sample under study was placed in a 4-cm-diameter beaker and sniffed at 25°C and at controlled relative humidity.