超高效液相色谱-静电场轨道阱高分辨质谱快速筛查和确证凉茶中167种非法添加药物

An accurate mass database and a method based on ultra high performance liquid chromatography-electrostatic field orbitrap high resolution mass spectrometry (UHPLC-Orbitrap HRMS) were developed. These were applied in the screening and identification of illegally added medicines in herbal tea. Based on investigations, 167 medicines were selected to build an accurate MS database; these medicines included antipyretic analgesics, glucocorticoids, antibiotics, and antihistamines, among other categories. The database was established using Orbitrap HRMS and TraceFinder software. The database carried information on all selected compounds, including the molecular formula, accurate mass of precursor ions and fragment ions, retention time, and mass spectra. The samples were ultrasonically extracted with a 50% (v/v) methanol aqueous solution. The extracted solutions were separated using a Waters XBrigde BEH C18 column (100 mm×2.1 mm, 2.5 μm). As the mobile phases, 0.1% (v/v) formic acid aqueous solution and acetonitrile containing 0.1% (v/v) formic acid were used, with gradient elution. The sample solutions were analyzed by Orbitrap HRMS in the full-scan MS and data-dependent MS/MS acquisition modes (Full MS/dd-MS(2)). Positive and negative polarity data were simultaneously acquired. Some parameters were optimized to increase the peak intensity and sensitivity of all compounds. The resolutions in the full-MS scan and dd-MS(2) scan were set to 70000 and 17500, respectively. In the full-MS mode, scanning was performed in the range of m/z 100 to 1000. In the MS/MS mode, the normalized collision energy (NEC) was set to 20%, 40%, and 60% for each compound. The inclusion list was not used during the measurement, and the dynamic exclusion time was set to 10.0 s. The loop count was set to 5. After acquiring the sample data with these conditions using Orbitrap MS, they were imported into TraceFinder software, through which the sample information was extracted and automatically matched with the information on compounds in the MS database. Screening and identification were conducted by comparing the retention times as well as the exact masses of precursor ions and fragment ions that were experimentally measured. If the errors between the experimentally and theoretically obtained masses of the precursor ions were below 5×10(-6) and the deviations in retention times were less than 20 s, then suspicious positive compounds might be identified. Furthermore, if such compounds possess more than one similar fragment ion with a mass tolerance below 5×10(-6), and exhibit similar ion distributions in the MS/MS profiles (compared to those in the database), they could be confirmed to be the same. The validation result showed that all compounds had good linear relationships, with correlation coefficients (r) greater than 0.99. Because pefloxacin, norfloxacin, desloratadine, astemizole and clindamycin had background interference, the method was not suitable for their quantification. Following experiments using three spiked concentrations, the recoveries of the rest 162 compounds were found to be in the range of 66.4%-118.1%, and the relative standard deviations (RSDs, n=6), in the range of 0.1%-16.1%. When the limit of detection (LOD) was 0.2 mg/kg, 83 compounds were detected, while when the LOD was 1.0 mg/kg, 167 compounds were detected. All compounds were matched successfully to the standard added sample with the MS database in TraceFinder software. To lower the likelihood of false positive and false negative results, a quality control method was recommended. The method was applied to analyze 245 herbal tea samples, among which 12 positive samples were detected. Thirteen positive compounds were found, including acetaminophen, diclofenac sodium, chlorpheniramine, brompheniramine, dexamethasone, dexamethasone 21-acetate, prednisone, prednisone 21-acetate, metronidazole, erythromycin, ciprofloxacin, amantadine, and dextromethorphan. In particular, amantadine, dextromethorphan, brompheniramine, and ciprofloxacin were newly detected, compared to standard methods. The developed method is rapid and accurate, and will be useful in the high-throughput screening of illegally added medicines in herbal tea.