[Simultaneous determination of 22 sunscreen agents in cosmetics by ultra-high-performance liquid chromatography using diode array detector].

A fast and simple ultra-high-performance liquid chromatography using diode array detector (UHPLC-DAD) technique has been developed for the simultaneous determination of 22 sunscreen agents (UV filters), viz. phenylbenzimidazole sulfonic acid (PBS), terephthalydene dicamphor sulfonic acid (TDS), benzophenone-4 (BZ4), camphor benzalkonium methosulfate (CBM), benzophenone-2 (BZ2), benzylidene camphor sulfonic acid (BCS), benzophenone-3 (BZ3), 3-benzylidene camphor (3BC), isoamyl p-methoxycinnamate (IMC), 4-methylbenzylidene camphor (MBC), diethylamino hydroxybenzoyl hexyl benzoate (DHHB), octocrylene (OCR), butyl methoxydibenzoyl methane (BDM), ethylhexyl dimethyl PABA (EDP), ethylhexyl methoxycinnamate (EMC), homosalate (HS), ethylhexyl salicylate (ES), diethylhexyl butamido triazone (DBT), ethylhexyl triazone (ET), drometrizole trisiloxane (DRT), methylene bis-benzotriazolyl tetramethylbutylphenol (MBP), and bis-ethylhexyloxyphenol methoxyphenyl triazine (EMT) in cosmetic products. Different parameters, such as column type, oven temperature, mobile phase composition, and detection wavelength, were studied. The best chromatographic separation was obtained under the following conditions: Poroshell 120 EC-C18 (100 mm×4.6 mm, 2.7 μm) column set at 25 ℃ and gradient acetonitrile-isopropanol-water (containing 50 mmol/L ammonium acetate and 0.05% (v/v) formic acid) as the mobile phase, pumped at a flow rate of 0.5 mL/min, with a wavelength of 311 nm. The proposed UHPLC-DAD technique provided separation of the 22 target sunscreen agents within 35 min, with the optimized sample pretreatment procedure below. First, samples were mixed thoroughly by adding 2 mL or 5 mL tetrahydrofuran, followed by vortex and dispersal. If the wax samples were still not homogenized completely, an ultrasonic dispersal protocol with heating to 50 ℃ was adopted. Second, the homogeneous samples were ultrasonically extracted with ethanol containing 0.1% (v/v) formic acid. The developed method showed good linear relationships, with correlation coefficients of no less than 0.998. Two kinds of samples with different matrix types were fortified at three levels. The average spiked recoveries of 22 UV filters ranged from 85.2% to 112.3%, with the relative standard deviations (RSDs) ranging from 0.5% to 6.5%. The limits of detection were between 0.7 and 64 mg/kg, and the limits of quantification ranged from 2.4 mg/kg to 215 mg/kg. Moreover, the stabilities of the mixed standard solutions at the levels of 2, 10, and 50 mg/L were tested. The stability results showed that drometrizole trisiloxane was stable for 12 h, while the others were stable for 36 h. The reliability of the developed method was demonstrated by applying it to 16 commercial sunscreen-containing cosmetic samples obtained from the Chinese market. The levels determined in this study agreed well with those of five commercial samples (such as emulsion and cream). The method developed was remarkably different from the standard method, which is mentioned in the Safety and Technical Standards for Cosmetics (2015 edition), especially in terms of mobile phase composition and extraction solvent. Compared to the standard method, this method bypassed the use of large amounts of corrosive solvents like tetrahydrofuran and perchloric acid, thus improving the extraction efficiency of low-polarity components like drometrizole trisiloxane, methylene bis-benzotriazolyl tetramethylbutylphenol, and bis-ethylhexyloxyphenol methoxyphenyl triazine, and the analytes were well separated with better stability. Benzophenone-2 was added to this method as another detection component. The good analytical features, as well as their environment-friendly characteristics, make the presented method suitable not only for routine analysis in cosmetics industries, but also as a candidate reference method for sunscreen analysis.