Abstract:

Adding glucocorticoids to cosmetics is strictly prohibited under Safety and Technical Specification for Cosmetics (2015) regulations. Accordingly, a method for simultaneously determining 100 glucocorticoids in cosmetics using ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) was developed. Apart from the 58 glucocorticoids listed in the two current national standards (GB/T 24800.2-2009 and GB/T 40145-2021) and the Safety and Technical Specification for Cosmetics (2015) regulations, 42 new glucocorticoids, including unreported metenolone acetate and dexamethasone 9,11-epoxide, were included in this method. Samples were dispersed in saturated sodium chloride containing 0.2% (v/v) acetic acid, and extracted with acetonitrile containing 0.2%(v/v) acetic acid. After that, an equal volume of water was added to the acetonitrile extraction solution to precipitate non-polar impurities. The solution was subsequently further purified using 10%(mass fraction) potassium ferrocyanide solution and 20%(mass fraction) zinc acetate solution containing 0.4% (v/v) acetic acid as precipitants. The sample solution was finally loaded onto a Waters ACQUITY UPLC BEH C₁₈ column (150 mm×3.0 mm, 1.7 μm) and gradient-eluted using aqueous and methanolic 0.2% (v/v) acetic acid solutions as mobile phases. The target compounds exhibited highly intense signals in positive electrospray-ionization mode. All compounds were analyzed in multiple reaction monitoring mode (MRM) and quantified using the external standard method. Rimexolone and hydrocortisone hemisuccinate hydrate exhibited peaks corresponding to their [M-H₂O+H]⁺ base ions, whereas the other 98 glucocorticoids exhibited [M+H]⁺ peaks. All positive samples were confirmed by their automated MRM-based information dependent acquisition (IDA) enhanced product ions (EPIs). Highly sensitive secondary mass spectra were acquired when EPI mode was combined with IDA and when target compounds in positive samples were located in MRM mode, with compounds further confirmed by matching their secondary mass spectra with those of the corresponding standard glycocorticoids using the Library Search function in SCIEX OS software.

Matrix effects (MEs) were investigated in water- and cream-based cosmetics by adding standard compounds to the extracted-matrix solutions. Ciclesonide exhibited no obvious matrix effect with a ME of 3.4% in the water-based matrix whereas it exhibited a serious inhibitory effect with a ME of 30.6% in the cream-based matrix. The remaining 99 target compounds exhibited MEs of -7.9%-12.4%, which are not significant because they are below 15%, further confirming that the sample-pretreatment method is effective and feasible. Therefore, ciclesonide in a cream-based matrix was quantified using a matrix-matched standard curve, while the others were quantified using solvent standard curves. The 100 glucocorticoids exhibited good linearities with correlation coefficients (r) of >0.99 in the range of 2.5-60 ng/mL. The limits of detection (LODs) and limits of quantification (LOQs) of this method were 0.03 and 0.1 μg/g, respectively. Recovery tests were performed at three spiked levels, namely one-, two-, and eight-times the LOQs. The 100 compounds exhibited recoveries of 79.6%-114.4% with relative standard deviations (RSDs) of 0.7%-9.4% (n=6) in water-based cosmetics; in comparison, they exhibited recoveries of 79.5%-112.1% with RSDs of 0.7%-12.9% (n=6) in cream-based cosmetics. Three cream-based positive samples were identified among 47 batches of cosmetics samples (including 25 batches of cream and lotion-based cosmetics, and 22 batches of water-based cosmetics), with quantitative analysis revealing five positive compounds: triamcinolone acetonide (115.1 μg/g), triamcinolone acetonide 21-acetate (0.6 μg/g), clobetasol propionate (24.5 μg/g), dexamethasone (2.7 μg/g), and dexamethasone-17-acetate (210.3 μg/g). These five compounds were further confirmed using MRM-IDA-EPI. The secondary mass spectra of the five compounds showed high matching degrees (>96%) with those of the corresponding standard glucocorticoids in the library database.

In conclusion, the method developed in this study is simple, practical, inexpensive, highly efficient, and suitable for qualitatively and quantitatively screening and analyzing 100 glucocorticoids in cosmetics, thereby filling supervisory gaps and greatly improving the detectability of glucocorticoids illegally added to cosmetics.

Key words: ultra performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS), glucocorticoid, cosmetic, illegally added