210 JOURNAL OF COSMETIC SCIENCE AVOBENZONE PHOTOSTABILITY IN SIMPLE POLAR AND NON-POLAR SOLVENT SYSTEMS C. Bonda, P. MarineIll, J. Trivedi, S. Hopper, G. Wentworth The C.P. Hall Company, Chicago, IL 60606 Introduction Avobenzone (Parsol© 1789, Roche) is the leading organic UV-A filter worldwide. Recent approval by the U.S. FDA likely means usage will increase substantially) Unfortunately, avobenzone's absorbance of UV radiation tends to decrease as its exposure to sunlight increases. 2 Researchers have identified two explanations for this tendency. The first is photolysis wherein UV radiation catalyzes fragmentation of the molecule? The second is UV-induced conversion of this [•-diketone from the enol to the keto tautomer (Figure 1). 4 In all known solutions, the enol predominates and is the sp•. ies responsible for absorption in the solar UV range. The keto tautomer absorbs UV below the range found in sunlight. 0 '•H 0 0 0 Enol Keto kmax = 350-355nm C20H2203 kmax = 260-265nm mw= 310.39 Fig. 1 Our research measured UV-induced tautomerization in the non-polar solvent cyclohexane when it has been modified by the addition of small amounts of other materials. We were especially interested in finding ff polar materials are more stabilizing than protic materials, or vice versa. In this way we sought to increase our understanding of the behavior of avobenzone in sunlight, and perhaps to find ways to mitigate its loss of absorbance in the solar UV range. Methods Avobenzone was procured from Roche. Octyldodecanol (Isofol © 20) and octanol (Alfol © 8) were procured from Condea-Vista, and butyloctyl salicylate (HallBrite TM BHB) was provided by C.P. Hall. The solvents and diethyl adipate were procured from Aldrich. The •H NMR studies were performed on a Bruker AM-400 spectrophotometer. UV radiation (290-400 nm) was provided by a 16S Solar Simulator equipped with a WG 320 filter and PMA 2100 Automatic Dose Controller (Solar Light Co.). The standard radiation dose in these experiments was 35 MED (735 mJ/cm2). UV spectra were measured on a CECIL CE 3021 spectrophotometer (Buck Scientific). Experiments were applied to 10 ppm avobenzone solutions in the following diluents: neat isopropanol neat cyclohexane 99% cyclohexanedl% isopropanol 99% cyclohexaned 1% tetrahydrofuran 99% cyclohexaned 1% octyldodecanol 99% cyclohexaned 1% butyloctyl salicylate 99% cyclohexaned 1% diethyl adipate 99% cyclohexanedl% octan& Remits The enol-keto equilibrium in deuterated cyclohexane was determined by NMR to be approximately 97%- 3%. We observed evidence of UV-induced tautomerization from the enol to the keto form in all solutions (Figures 2, 3). After irradiation was stopped, we observed in all solutions that absorbance of the keto immediately started to decline and, correspondingly, absorbance of the enol immediately started to rise (Figure 4). This process continued until a new equilibrium was reached. There was some, apparently permanent reduction in total avobenzone concentration. As $nmmarized in Table 1, hydroxylic (and, therefore, protic) additives such as primary alcohols and butyloctyl salicylate were more succes,nful at inhibiting enol-keto tautomerization than were aprotic polar additives and non-polar additives which were also aprotic.

PREPRINTS OF THE 1998 ANNUAL SCIENTIFIC SEMINAR 211 Rank IMluent % OHglnal Absorbance after Molecular Weight of 35 MED ((• 353nm} diluent additive I 99% cyclohexane./1% IPA 91.2% 68 2 Neat isopropanol 90.9% -- 3 99% cyclohexaneYl % octanol 88.7% 130 4 99% cyclohexaneYl % butyloctyl salicylat• 84.5% 306 5 99% cyclohexane./1% octyldodecanol 78.9% 298 6 99% cyclohexaned 1% tetrahydrofuran 69.2% 72 7 Neat cyclohexane 61.8% -- 8 99% cyclohexaneY1% diethyl adipate 58.8% 174 Table I Conclusions The tendency of avobenzone to lose absorbance in sunlight is explained in part by' UV-induced enol to keto tautomerization. Surprisingly, the tautomerization reverses when irradiation ceases. and something resembling the original enol-keto equilibrium results. Stability during [IV irradiation of the enol-keto equilibrium is increased in what appears to be a concentration-dependent manner by adding hydroxylic (protic) materials to the solvent. Butyloctyl salicylate increases stability of the enol-keto equilibrium disproportionately when compared to the nearly equimolar octyldodecanol. Polar materials which are not hydroxylic do not appreciably improve stability of the equilibrium. Our hope is that these findings will lead to improvements in avobenzone formulations and, therefore, to more protective, dependable, and efficacious sunscreens. References 1. Food and Drag Administration, Announcement of Enforcement Policy, Fed. Register, Vol 62, No. 83, April 30, 1997 2. C. Bonda, Formulating Stable, High SPF, Broad Spectrmn Sunscreens with Avobenzone, from poster presented at SCC Annual Scientific Meeting, NYC, (1997) 3. W. Schwack and T. Rudolph, Photochemistry of Dibenzoyl Methane UVA Filters Part I, J. Photochem. & Photobiol., 28, 229-234, (1995) 4. L. Andrae et al, A UVA Filter (4-tert-butyl-4'-methoxydibenzoylmethane): Photoprotection Reflects Photophysical Properties, J. Photochem. & Photobiol., 37, 147-150, (1997) Acknowledgements The authors gratefully acknowledge the valuable assistance and contibufions of Yin Hessefort, Magda Roth, Urvil Shah, and Robert McMillin of the C.P. Hall Company, Dr. John Harwood of the University of Illinois-Chicago, and David Steinberg of Steinberg & Associates.