[Determination of 32 oxidative dyes by high performance liquid chromatography and confirmation by high performance liquid chromatography-tandem mass spectrometry]
Zhiming Zhou,Jing Li,Zhanghao Chen,Yingying Wu,Tuliu Lu,Shuxiong Xiao
DOI: https://doi.org/10.3724/SP.J.1123.2022.03003
Abstract:Various types of oxidative dyes used in hair dye products possess poor stability and have varying frequency of use. Interference problems also frequently arise in actual sample measurements. Therefore, it is necessary to establish a simple, rapid, accurate, and specific method for the determination of common dyes in hair dye products for their effective regulation. In this study, dyes were grouped according to their frequency of use. Using a C18 column that minimizes the silanol effect and influence of metals, the quantitative high performance liquid chromatography (HPLC) method for 32 dyes listed in Safety and Technical Standards for Cosmetics (2015 edition) was optimized, and a high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) confirmatory method for the dyes was established. The samples were extracted using a mixed solution of ethanol-water (1∶1, v/v) with 10 g/L sodium bisulfite solution as an antioxidant, vortexed and mixed, and then extracted by ultrasonication in an ice bath for 10 min. Methanol, acetonitrile, and phosphate buffer were used as the mobile phases in the HPLC analysis. Additionally, two different elution conditions (chromatographic gradient) were used for the separation of 32 oxidative dyes, which were detected at a wavelength of 280 nm. The HPLC separations were compared using columns of particle sizes 5 μm and 2.7 μm; 5 μm C18 columns with better anti-interference and antiblocking ability were selected. Satisfactory separation was achieved for all three commercial C18 columns with a particle size of 5 μm, and the method had good general usability. In condition 1, 17 commonly used dyes and three less commonly used dyes were assigned to group Ⅰ and separated by HPLC; in condition 2, eight banned dyes and four other less commonly used dyes were assigned to group Ⅱ and separated by HPLC. The HPLC-MS/MS method used 5 mmol/L ammonium acetate aqueous solution-acetonitrile and 5 mmol/L acetic acid aqueous solution-acetonitrile as mobile phases in the positive and negative ion modes, respectively. Multiple reaction monitoring (MRM) was performed for qualitative and quantitative analyses in the electrospray ionization mode. Under the examined conditions, six pairs of isomers were well resolved. For the HPLC and HPLC-MS/MS methods, the relative standard deviations (RSDs) of the intra-day precision and 48 h stability tests were less than 10%. The recoveries were between 82.6% and 114.9% (RSD<10%). In the HPLC method, 32 dyes showed good linearity in an approximate range of 10-500 mg/L (r2>0.99), and the limits of detection (LODs) were 9.7-40.1 μg/g. The linear range of hydroquinone in the HPLC-MS/MS method was 2.0-79.7 mg/L, and the LOD was 8.0 μg/g; the linear ranges of the other components were approximately 0.1-4 mg/L, and the LODs were 0.01-0.4 μg/g. The actual samples were simultaneously measured by HPLC, HPLC-MS/MS, and the standard method. Finally, 16 of the 32 dyes were detected, and the detected contents ranged from 58 to 25160 μg/g. The RSDs of the results obtained from the three detection methods were between 1.9% and 10.1%. All detected components were within the limits of group Ⅰ of this method. In comparison with methods reported in the literature and the standard method, this method covers all components for the routine regulatory inspection of oxidative dyes in cosmetics. The method can separate the commonly used dyes under the same HPLC conditions and avoid interference from 15 other commonly used dyes during the analysis of actual samples. A suitable HPLC-MS/MS confirmatory method was also established for the identification of currently unknown substances in the statutory inspection of cosmetics. The method is simple, rapid, accurate, and specific with general usability and good operability.