Determination of 2 Preservatives in 3 Common Rinse-Off Cosmetics by UHPLC-MS/MS

A method for determination of methylisothiazolinone (MI) and methylchloroisothiazolinone (MCI) in rinse-off cosmetics by UHPLC-MS/MS was established. The sample (0.2 g) was ultrasonically extracted for 10 min with 10% (volume fraction) methanol solution, then 100 g·L^-1 potassium ferrocyanide solution (0.2 mL) and 200 g·L^-1 zinc acetate solution (0.2 mL) were used to precipitate large molecular impurities such as protein in the sample. The supernatant obtained after settling for 5 min was filtered through a 0.22 μm filtering membrane, and the two target compounds were eluted on a Waters ACQUITY UPLC HSS T3 column by mixed solutions composed of methanol and 0.1% (volume fraction, the same below) formic acid solution with different volume ratios. For the first 4.5 min in the elution program, only 0.1% formic acid solution was used to remove water-soluble impurities such as surfactants. The separated targets were ionized in ESI⁺ mode, and MRM was adopted in MS/MS for detection. The results showed that linear relationships between values of peak area and mass concentration of MI and MCI were found in the ranges of 0.8-200 μg·L^-1 and 1.6-400 μg·L^-1, with detection limits (3S/N) of 0.02, 0.01 μg·g^-1, respectively. The spiked recovery tests were made on the blank samples, giving values of recovery in the ranges of 94.3%-107% and 93.4%-106%. RSDs of the determined values obtained from 6 repeated test on actual samples were 0.70% and 2.1%. The method was applied to the analysis of 90 batches of bath, hair washing and facial cleansing cosmetics:the content of MI in one batch of baby and child bath product was close to the limit value (0.01%) of Safety and Technical Standards for Cosmetics; the content of MI in 7 batches of bath and 4 batches of hair washing products exceeded the limit value (0.001 5%) of the European cosmetics regulation EC 1223/2009; the content of MI in 8 batches of products was above 5 times more than that of MCI, and it was suspected that MI and kathon were added to the sample at the same time, which did not comply with the provision of Safety and Technical Standards for Cosmetics; in the 90 batches of products, compositions of 25 batches of products were found to be inconsistent between the identification of the label and the actual detection.