128 JOURNAL OF COSMETIC SCIENCE (20). Estimates of daily human exposure to DEP, based on ester metabolite concentra- tions in urine, range from undetectable to 320 mg/kg/day (21). Phthalate esters, are of potential concern because of reproductive and other toxic effects reported for phthalate esters in animal models. Tests with rats have shown that some phthalate esters can damage the male reproductive system of offspring and cause other developmental abnormalities (22-36). The relevance of human exposure to phthalate esters is difficult to assess because the effects observed in animals resulted from exposure to relatively high doses (on the order of 100 to 1000 mg/kg/day), and more importantly, rats may metabolize phthalate esters differently than humans (3 7 ,38). While one study using a comet assay suggested that the urinary metabolite of DEP at levels currently found in the environment might cause human sperm DNA damage (39), another study showed primates were less sensitive to phthalate exposure than rodents (40). Further, because of their ubiquitous use, it is difficult to pinpoint any specific source of phthalate ester uses as being potentially responsible for any observed effects. In July 2002, the Environmental Working Group (EWG), a coalition of environmental and public health organizations, reported on the analysis of 72 cosmetic products for the phthalate esters DMP, DEP, BBP, DBP, and DEHP (15). According to the report, phthalate esters were found in 52 of the 72 products tested, at levels ranging from less than 50 parts per million to nearly three percent. Also, according to the report, none of the 52 products listed the phthalate esters as ingredients on the product labels. Based on the safety concerns raised by the EWG and other published data, the FDA initiated a project to determine consumer exposure to phthalate esters from cosmetic products (41). The present study was undertaken to develop and validate an analytical method for phthalate esters in cosmetic products, to verify the levels reported in cosmetics by the EWG, and to collect additional data on phthalate ester levels in other types of cosmetic products. A large variety of analytical methods have been published for the analysis of consumer products, biological materials, and environmental samples for phthalate esters. Methods based on gas chromatography with either flame ionization or electron capture detection have been described for food simulants (42), water (43), plasma (44) and edible oils (45). High-performance liquid chromatography (HPLC) with UV detection has been applied to water (46), IV drug solutions (47), blood plasma (48), and water (49). More recently reported methods have utilized mass spectrometry for the detection of phthalate esters. Methods applying gas chromatography coupled to a mass spectrometer have been de- scribed for the analysis of water (50-53), saliva (54), and plastic and PVC materials (55,56). HPLC and mass spectrometry have been used for the analysis of urine (57-59), human milk (20), and IV drug solutions (60). Very little has been published on ana- lytical methods for the determination of phthalate esters in cosmetic products. Two older methods, based on simple gas chromatography, have been described for cosmetic prod- ucts ( 61,62), but little data has been provided. The present study describes a method for the determination of phthalate esters in a variety of different cosmetic products. The method utilized a Celite column extraction method originally developed in our labora- tory for the analysis of phenol, resorcinol, salicylic acid, and a-hydroxy acids in cosmetic products and salon preparations (63). The method was validated for the determination of phthalate esters in several different types of cosmetic products.

PHTHALATE ESTERS IN COSMETIC PRODUCTS 129 EXPERIMENT AL REAGENTS AND MATERIALS The following reagents and materials were used: Hexane was purchased from Burdick & Jackson (Muskegon, Michigan). Acetonitrile and methanol were purchased from T. J. Baker (Phillipsburg, New Jersey). 2-Propanol was purchased from Fisher Scientific (Fairlawn, New Jersey). All solvents were HPLC grade. Phthalate esters DMP (99%), DEP (99.5%), and BBP (98%) were purchased from Sigma Aldrich (Milwaukee, Wis- consin). DBP (2:::98%) and DEHP (99.5%) were purchased from Sigma Aldrich (St. Louis, Missouri). De-ionized water was prepared with a Milli-Q purification system from Millipore (Billerica, Massachusetts). Celite 545 was purchased from Fisher Scientific (Fairlawn, New Jersey). The extraction tubes and filter disks were obtained from Supelco (Bellefonte, Pennsylvania). PHTHALATE ESTER CALIBRATION STANDARDS A primary standard solution of a mixture of the five phthalate esters (- mg/ml each) was prepared by adding approximately 100 mg of each to a 100-ml amber volumetric flask and diluting to the mark with hexane. Because of the wide range of possible concen- trations in cosmetic products, three sets of working standards were prepared. One set was prepared at approximately 0.001, 0.003, 0.006, and 0.01 mg/ml by appropriate serial dilution of the stock solution. Similarly, a second set was prepared at approximately 0.01, 0.03, 0.06, and 0.1 mg/ml, and a third set for BBP, DBP, and DEHP only was prepared at 0.10, 0.30, 0.60, and 1.00 mg/ml. HPLC peak areas were determined based on duplicate injections of 20 µl, and a calibration curve was obtained by plotting peak area versus standard concentration. SAMPLE EXTRACTION To avoid contamination by environmental sources of phthalate esters, all glassware was thoroughly cleaned and rinsed with water and ethanol before use, and phthalate- containing plastics were avoided. Approximately 1 g of each cosmetic sample was weighed into a 40-ml beaker, mixed thoroughly with about 3 g of Celite, and then transferred to a 15-ml extraction tube containing a filter disk. The sample/Celite mix- ture was covered with a second filter disk and compacted firmly with a stirring rod. The prepared column was eluted with sufficient hexane to obtain 10 ml of extract rn a volumetric flask. The extraction flask was mixed well prior to HPLC analysis. HPLC ANALYSES HPLC analyses were carried out on an Agilent ll00 series HPLC, equipped with a quaternary pumping system, an in-line vacuum degasser, a variable wavelength diode array UV-visible absorbance detector, a 20-µl injection loop, and a personal computer with HP Chemstation software. Chromatographic separation was achieved using a Whatman Partisil ODS-3 5-µm guard column (7 .5 mm by 4.6 mm ID) and a Whatman Partisil ODS-3 5-µm analytical