From reaction energetics to automated analysis tool auxiliary identification: A self-validating multidimensional liquid chromatography- mass spectrometry framework for characterizing polymerized impurities in carbapenem antibiotics.

The presence of polymerized impurities in carbapenem antibiotics poses a public health risk due to their potential to induce passive cutaneous allergic reactions. Previous literature has shown that isomeric impurities can exhibit different biological properties and toxicological profiles. Therefore, studying the polymerized and isomeric impurities of carbapenem antibiotics is highly significant. However, these impurities have received relatively little attention in previous studies. Conversely, current studies on the structural elucidation of pharmaceutical impurities often neglect to investigate formation mechanisms. They also neglect the in-depth analysis of MS/MS fragment ion disparities, which can result in incorrect structural attribution. In order to characterize impurities and elucidate polymerization mechanisms, this study has developed an integrated analytical strategy incorporating multidimensional chromatographic separation, high-resolution mass spectrometry, theoretical calculations, and an automated analytical tool assisted in the auxiliary identification. HPSEC and RP-HPLC methods with novel separation principles and two-dimensional chromatographic methods (UHPLC-Q-TOF MS and 2D HPSEC × LC-Q-Exactive MS) were established to resolve polymerized species in biapenem and meropenem. This enabled the systematic separation and identification of 15 impurities (including 11 polymerized variants) through multistage MSⁿ fragment analysis in dual ionization modes. On this basis, the possible dimerization mechanism of the dimeric isomers was elucidated via theoretical calculations. This allowed both the dominant polymerization sites and the dimerization reaction to be recognized. Additionally, our self-developed automated analytical tool assisted in the auxiliary identification of dimeric isomer configurations in actual samples. The dimer configurations identified from the LC-MS/MS results were consistent with the dominant dimer configurations revealed by theoretical calculations. This work establishes a self-validating methodology that integrates "multidimensional separation-high-resolution MSn characterization -mechanistic validation- automated analysis tool auxiliary identification analysis," for the precise characterization of polymerized impurities.