144 JOURNAL OF COSMETIC SCIENCE returned to the values noted before irradiation (OD at 358 nm: 0.15). The UVA absorbance decrease was slower for the preparation with the Cl0-DBM than for that of the preparation with BM-DBM. The triangles in Figure 8 (corresponding to Ml in Table I) plot the variation of the absorbance of the cream containing 1 % w/w of a BM-DBM/ClO-DBM mixture (molar ratio 7:3). Absorbance at 358 nm decreased progressively. After five, ten, and 20 minutes of irradiation, the cream had lost, respectively, 45%, 65%, and 80% of its absorbance in the UVA (compared to 50%, 80%, and 96% for the preparation with BM-DBM alone). The rate of absorbance diminution was slower than that of the preparation containing BM-DBM. BEHAVIOR UNDER IRRADIATION WITH NATURAL SUNLIGHT The same preparations were exposed to solar light. Each cream was tested under the conditions described above (see Table II). Figure 9 shows the variations of absorbance of creams with 1 % w/w Cl0-DBM versus time of exposure to natural sunlight. Two assays (C' 1 and C'2 in Table II) were carried out and confirmed the reproducibility of the method. Figure 10 shows the variation of absorbance at 358 nm of preparations con­ taining 1 % w/w of filter under natural sunlight. We observed a strong decrease in the absorbance of the cream contarnrng 1 % w/w BM-DBM (diamonds in Figure 10, corresponding to P' 1 in Table II). Indeed, after one hour, two hours and three hours of irradiation, the cream lost 64%, 85%, and 92%, respectively, of its UV A absorbance. 1,2 1 E C: CX) 0,8 � 0,6 C: .0 0,4 .0 0,2 0 0 20 40 60 80 100 120 140 160 180 Time (minutes) Figure 9. Absorbance of preparations containing 1 % w/w ClO-DBM under natural sunlight irradiation (c'l: triangles C'2: stars).

NEW LONG-CHAIN UV ABSORBER 145 1,2 1 E C: a:, 0,8 m � 0,6 .c 0,4 .c 0,2 0 0 20 40 60 80 100 120 140 160 180 Time (minutes) Figure 10. Absorbance of preparations containing 1% w/w BM-DBM (P'l: diamonds) or ClO-DBM (C'l: squares) or a BM-DBM/ClO-DBM mixture, 7:3 molar ratio (M'l: triangles) under natural sunlight. Initially, the cream containing 1 % w/w ClO-DBM (squares in Figure 10, corresponding to C'l in Table II) presented very low absorbance in UVA (OD at 358 nm: 0.13). Like under xenon lamp irradiation, UVA absorbance increased as irradiation started. After 60 minutes of irradiation, an increase of 58% in the absorbance at 358 nm was noted. The absorbance reached a maximum (OD at 358 nm: 0.27) and remained stable for 30 minutes. Then it began to decrease slowly. After two hours of irradiation, we observed a decrease of 3% in comparison with the maximum absorbance obtained (85% for BM-DBM). After three hours, we noted a decrease of 22% (92% for BM-DBM) in comparison with the absorbance at one hour of irradiation. The absorbance was always above the initial one (OD at 358 nm: 0.21). Concerning the cream containing 1 % w/w of a BM-DBM/ClO-DBM mixture (molar ratio 7:3) (triangles in Figure 10, corresponding to M' 1 in Table II), the UV A absor­ bance at 358 nm remained constant during the first 30 minutes of solar exposure. Then it decreased slowly and progressively. After one hour, two hours, and three hours of irradiation, the cream lost 11 %, 16%, and 34%, respectively, of its absorbance in UVA. The rate of absorbance diminution was slower than that of the cream containing only BM-DBM. DISCUSSION Beta-diketone compounds exist as keto/enol tautomers. Considerable attention has been focused on the equilibrium populations of each tautomer in various solvents, especially for dibenzoylmethane and its derivatives (16,22-24). The keto/enol tautomer ratio depends on the nature of the a-substituents, the nature of the solvents, the temperature,