QUANTITATIVE ANALYSIS OF BERGAPTEN 251 perfume applied to filter paper discs and also to silica gel. Yves Saint Laurent (YSL) toilet water (for males) was chosen as an example. Aliquots of 5 /xl were chromatogrammed as before and a similar volume was added to 10 mm Whatman No. ! filter disc and also to TLC plate. A total of ten agar plates were inoculated with Co al- bicans and to each plate a 10 mm disc, silica gel containing bergapten TLC fraction and silica gel containing the whole perfume (not chromatogrammed), was added. Five plates were exposed to UVA and the rest were kept in the dark. The killing zones were measured and the 'results subjected to statistical analysis. In addition to these experi- ments, all perfumes were tested for direct toxicity to the yeast by pipetting 25/xl of each perfume to 10 mm filter paper discs and dried under a stream of warm air from a hair dryer. The discs were added to seeded agar plates which were then grown in the dark. Any inhibition of the yeast growth around the discs was taken as a sign of toxicity. The toxic effects of each perfume were later compared with its phototoxic effects by growing one plate under UVA illumination and a duplicate in the dark. THE SPECTROPHOTOMETRIC TEST The technique used is a modification of the method describe•l by Cieri (6) and later used for the estimation of the 5-MOP content of bergamot oil (2). In this technique, sixteen 25/xl aliquots of each perfume and a similar number of 25/xl aliquots of 1% bergamot oil in ethanol along with 25 /xl of 0.05% pure 5-MOP in ethanol used as reference were chromatogrammed as previously described. The silica gel containing the 5-MOP fractions was scraped from the plates onto clean paper. All 5-MOP frac- tions of bergamot oil were transferred to a glass-stoppered tube and those of the perfume combined in another tube. 5-MOP was then extracted in 10 ml chloroform, centrifuged at 2000 rpm for 5 min and the absorption spectrum of the clear supernate obtained using an SP 8000 recording spectrophotometer (Unicam) with chloroform in the reference cell. The absorption at 310 nm was recorded and the concentration of bergamot oil and 5-MOP in each perfume was calculated as follows: 1. Percentage bergamot oil (v/v) = perfume absorption/bergamot oil absorption where perfume absorption = absorption of supernate containing bergapten TLC frac- tion of perfume, and bergamot oil absorption = absorption of supernate containing bergapten TLC fraction of bergamot oil. 0.27 2. Percentage 5-MOP (w/v) = Percentage bergamot oil x -- 100 RESULTS THE CANDIDA ALBICANS PHOTOTOXICITY TEST TLC Patterns. On TLC, bergamot oil showed five main fluorescent bands (Figure 1). Band 1, 2 and 3 are known as bergamottin, 7-geranoxycoumarin and citropene respec- tively, while band 4, which has a yellowish-green fluorescence, is 5-MOP (2,6). The TLC pattern of perfumes containing bergamot oil was variable and for the sake of sim- plicity these perfumes were divided into two groups. Group A (e.g., YSL toilet water) had five and sometimes six main fluorescent bands similar to those of bergamot oil

252 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS BERGAMOT OIL YVES ST. LAURENT TOILET WATER Solvent front .., Origin 5 Figure 1. Thin layer chromatogram showing 1% bergamot oil in ethanol and Yves Saint Laurent toilet water as observed under UVA illumination. The perfume shows six main fractions, five of which are similar to those of bergamot oil. Band 4 is 5-MOP (Figure 1). The bergapten TLC band of these perfumes was adequately separated from the adjacent TLC bands and had a yellowish-green fluorescence similar to that of the 5-MOP fraction of bergamot oil. However, the fluorescence of the 5-MOP band of this group of perfumes and of bergamot oil was not as bright and clear as that of pure 5-MOP. Perfumes belonging to Group B (e.g., Brut aftershave), which formed the larger portion of the perfumes tested, had more than five to six main fractions and the 5-MOP band of most of these perfumes did not show a good separation from the ad- jacent bands and had a dull yellowish fluorescence indicating the presence of im- purities. This was further confirmed by the absorption spectrum which was different from that of pure 5-MOP (Figure 2). Steam distillation, which was used in Cieri's method, was carried out to separate the steam-volatile portion of the perfume and