This study aimed to select and optimize a sample extraction method for the determination of nitrite from milk powder. The absorbance of the experimental water stored for different periods of time was detected by three extraction methods, heat extraction, ultrasonic extraction and ultrasonic decolorization. The color and spiked recoveries of the filtrates from the three methods were evaluated. The nitrite content of quality control samples was determined by ultrasonic decolorization and the limit of detection (LOD) of this method was investigated. The results indicated that storage time did not affect the absorbance of the experimental water. The recoveries of the ultrasonic decolorization method were over 90%, with relative standard deviations (RSD) less than 10%. and the measured results of the quality control samples were within the characteristic value range with LOD of 0.4 mg/kg, which meets the laboratory quality control requirements. The ultrasonic decolorization method has good accuracy and accuracy, and it is recommended that it be used for the decolorization of colored substances in the national standard method.

Using gas chromatography-tandem mass spectrometry (GC-MS), an indirect method was established for the simultaneous detection of 3-chloropropanol esters (3-MCPDE) and glycidyl esters (GE) in infant formula for special medical purposes. The conditions for sample extraction and transesterification/hydrolysis were optimized and stable isotope dilution was used to improve the accuracy and precision of the method. The results showed that the method exhibited good linearity in the range of 0.005–2.000 μg/mL (R2 ≥ 0.999), together with a limit of detection (LOD) of 5 μg/kg and a limit of quantification (LOQ) of 25 μg/kg (both calculated based on the corresponding amount of chloropropanol). The recoveries of the analytes spiked in infant formula for special medical purposes at 25, 125 and 250 μg/kg were 88.5%–104.1%, and the relative standard deviations (RSDs) for six replication determinations were less than 15% (n = 6). The levels of 3-MCPDE and GE in three types of commercial infant formula for special medical purposes were below the maximum limits set by the EU.

A high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method was developed for the determination of vancomycin, norvancomycin and teicoplanin in milk samples. Selective reaction monitoring was performed and the external standard method was used for quantification. Among various solvents tested, 0.1% formic acid-acetonitrile (85:15, V/V) was selected. Samples were ultrasonically extracted and vancomycin and norvancomycin were cleaned up by cation exchange column chromatography, while teicoplanin was purified using a C18 solid-phase extraction (SPE) cartridge. Acetonitrile-0.1% formic acid solution was chosen as mobile phase for gradient elution. The analytes were separated on a reversed-phase C18 column before being analyzed by mass spectrometry. The limits of detection (LOD) for vancomycin, norvancomycin and teicoplanin were 2, 1 and 2 μg/kg, and the limits of quantitation (LOQ) were 4, 2 and 4 μg/kg, respectively. The recoveries of the method ranged from 77.3% to 84.5% at spiked levels equal to LOQ, 5 × LOQ and 10 × LOQ, with relative standard deviations (RSD) of 4.7%–7.2%. The method could be used in the qualitative and quantitative detection of vancomycin, norvancomycin and teicoplanin in milk samples.