1532 SHORT PAPERS Analyst, Vol. 108 Isolatian of Bergapten and Limettin from Bergamot Oil Vincent B. Croud, John R. Michaelis and Arnold G. Pindar Labovatory of the Government Chemist, Cornwall House, Stamford Street, London, SE1 9NQ Keywords ; Bergapten ; limettin ; lactonic fraction ; thin-layer chromatography ; bergamot oil Suntan products and some perfumes may contain citrus oils in order to promote tanning or enhance fragrance. These oils contain coumarins, which may have a phototoxic effect. Probably the best known of these is 4-methoxyfuro[3,2-g]chromen-7-one (bergapten) (I), also commonly known as 5-methoxypsoralen (&MOP), which is present at a level of 0.37% in bergamot oi1.l Marzulli and Maibachl in a detailed study of the phototoxicity of bergamot oil concluded that Berlock dermatitis, a condition similar to acute sunburn, can be caused by the application of products containing bergapten and exposure to sunlight. The phototoxic effects of other furocoumarin components are less certain. However, the isolation of 5,7- dimethoxycoumarin (limettin) (11) , to which bergapten may be converted, has been included in this study.This laboratory was asked to examine a number of suntan products and a perfume sample for the presence of bergapten and limettin. The determination was to be carried out by thin- layer (TLC) and high-performance liquid chromatography (HPLC) for which standard samples were required for comparison purposes. No such materials were directly available at that time and therefore it was decided to extract and purify the materials from readily available bergamot oil.A literature search yielded many methods for the determination of coumarins and psoralens in citrus oils, of which four pieces of work were particularly r e l e ~ a n t . ~ - ~ The procedures described can be separated into two distinct stages: (i) the isolation of the lactonic species from the oil; (ii) the extraction of the pure material from this fraction. Crown Copyright.December, 1983 SHORT PAPERS 1533 Considering stage (i), Suzuki et aL2 suggested an initial clean-up by eluting the sample through a Florisil column with ethyl acetate, followed by a second stage with a Florisil column and eluting with methylene chloride. The main problem in stage (ii) was the separation of bergapten from limettin.The final yield was poor and the method was, therefore, rejected. I CH3 II Sethna and Shah3 proposed a more rigorous treatment, which in fact Suzuki et aL2 had applied to some samples. This involved base hydrolysis with methanolic potassium hydroxide in order to open the lactone ring and produce hydrophilic compounds. After removal of the other components of the oil by ether extraction, the lactone ring is closed by warming with 10% sulphuric acid and the bergapten is extracted with chloroform. Both Wisneski4 and Cieri5 suggested the removal of any volatile fractions from the oil as an initial clean-up procedure, prior to isolating the lactonic fraction. The former used a vacuum distillation, the latter a steam distillation. Wisneski followed the vacuum distillation with the procedure of Sethna and Shah3 to isolate the lactonic fraction.Ether was then added to this fraction, it was cooled and crystals of bergapten and limettin were obtained that were then recrystallised four times from hot methanol. However, the authors found that the bergapten and limettin were still contaminated by other materials. It was found that elution from a silica gel 60 column with hexane - ethyl acetate (75 + 25 V / V ) led to the isola- tion of several fractions that contained solely bergapten and limettin. The first few fractions contained neither bergapten nor limettin, and these were followed by four fractions in which limettin only was found. Finally about ten fractions containing both bergapten and limettin were eluted. In this work the following isolation procedure was employed: (i) removal of the volatile fraction by vacuum distillation (this was preferred to the steam distillation as it was found easier to control and was less likely to decompose the sample); (ii) isolation of the lactonic species using the procedure of Sethna and Shah; and the separation of bergapten and limettin by TLC.During the course of the work it was noted that bergapten slowly decomposed to limettin in the presence of light, so wherever possible the apparatus was protected from light. Some bergapten was obtained from an alternative source; this was used to confirm the identity of the product obtained from this procedure. This bergapten was also found to contain a trace of limettin. Other possible procedures to clean up the lactonic fraction were sought.Procedure Take bergamot oil (50 ml or 45 g) and distil up to 60 "C under vacuum (1-2 mmHg) (1.3- 2.6 mbar). Stir and reflux the residue for 3 h with a mixture of potassium hydroxide (10 g) in methanol (67 ml) and water (33 ml). Cool and add saturated sodium chloride solution (100 ml), transfer into a separating funnel, wash with ether (2 x 50 ml) and discard the ether washings. Neutralise the solution with sulphuric acid (10% V/V) and add a further portion of acid (30 ml). Heat on a water-bath for 1 h. Extract with chloroform (3 x 50 ml) and wash successively with 1% (m/V) sodium hydrogen carbonate solution (2 x 25 ml) and water (25 ml). Dry the chloroform extract with anhydrous sodium sulphate and filter. Add xylene (10 ml) and remove the chloroform on a rotary evaporator.1534 SHORT PAPERS Analyst, V d .108 Condition a silica gel 60 column (150 mm x 15 mm i.d.) with hexane - ethyl acetate (75 + 25 V/V). Add the residual xylene solution from the rotary evaporator to the column and elute with the same mobile phase that was used to condition the column. Collect 10-ml fractions. Check each for the required substances by TLC on silica gel 60 using hexane - ethyl acetate (75 + 25 V / V ) as the mobile phase. (Bergapten gives a strong green fluorescence at an R, of 0.20 when irradiated at 366nm and limettin shows a strong blue fluorescence at an R, of 0.23.) Combine the fractions containing bergapten or limettin, evaporate to dryness in a rotary evaporator and redissolve in acetone (5 ml).Apply an aliquot (100 pl) of this solution to a silica gel 60 TLC plate (200 mm x 200 mm, silica 0.25 mm thick) as a band of 180 mm in length. Remove the band of the required material (bergapten or limettin) from the plate and wash with acetone. Filter and remove the acetone in a rotary evaporator to yield either bergapten or limettin. Elute the plate four times with hexane - ethyl acetate (75 + 25 V / V ) . Results and Discussion Both bergapten and limettin were recovered at a level of 3 mg per plate. The literature1 gives the concentration of bergapten in bergamot oil as 0.37%, therefore 92% of the bergapten was recovered. As bergamot oil is a natural product the concentration of bergapten may vary depending on the source of the oil, so 92% is purely a guide to the efficiency of the procedure.TABLE I ANALYSIS OF THE PURITY OF LIMETTIN AND BERGAPTEN BY ULTRAVIOLET SPECTROSCOPY AND MASS SPECTROMETRY Mass spectrometry I Ultraviolet spectroscopy A , \I Amax. (limettin)/ Amax. (bergapten)/ ' nm nm mlz 219 219 206 241* 248 178 321 257 163 266 149 312 135 85 * Shoulder on peak. Limettin A 7 Relative Molecular intensity ion 100 C11H1004 82.7 C,,H,,O, 48.1 C,H,O, 16.4 25.0 18.3 I 1mlz 216 201 188 173 145 89 1 Bergapten v --A- Relative Molecular intensity ion 100 C12H804 30.9 C11H504 12.1 C,,H8O, 66.2 C,,H,O, 30.4 C,H50, 23.2 C,H5 The purity of these compounds was checked by both ultraviolet spectroscopy and mass spectrometry and the results are presented in Table I. The mass spectrometric results showed that limettin was absent from the bergapten and vice veysa. The solution of bergapten and limettin in acetone obtained from the silica column was retained for six weeks in order to obtain further supplies of the purified materials. In this period the recovery of bergapten decreased to about 2 mg per plate; however, this was still an adequate amount for use as a standard. The solution was stored in a darkened bottle in a refrigerator in order to slow down this decomposition. This work forms part of a larger programme of work undertaken by the Laboratory of the Government Chemist on behalf of the Department of Trade. References 1. 2. 3. 4. 5. Marzulli, F. N., and Maibach, M. D., J. SOC. Cosvnet. Chem., 1970, 21, 695. Suzuki, H., Wakamura, K., and Iwaida, M., J. SOG. Cosmet. Chem., 1979, 30, 393. Sethna, S. M., and Shah, M. N., Chem. Rev., 1945, 36, 1. Wisneski, H., J . Assoc. Off. Anal. Chem., 1976, 59, 547. Cieri, U. R., J. Assoc. 08. Anal. Chem., 1969, 52, 719. ' Received May 26th, 1983 Accepted July 28th, 1983