Comprehensive approach for dissolved organic matter chemical characterization using orbitrap fusion tribrid mass spectrometer coupled with ion and liquid chromatography techniques

Dissolved organic matter (DOM) contains the largest active organic carbon in the global carbon cycle. Though it has been extensively studied, only <10% of DOM has been chemically characterized into individual dissolved compounds. DOM is one of the most complex mixtures in nature; it consists of thousands of compounds with various molecular compositions, functional groups, and physicochemical characteristics. This study introduced a more comprehensive DOM characterization method by coupling both Ion Chromatography and Liquid Chromatography with high accuracy mass and resolution mass spectrometer. Estuarine DOM samples were and analyzed by Orbitrap Fusion Tribrid mass spectrometer (OT-FT-MS) coupled to both Ion Chromatography (IC) in negative mode and Liquid Chromatography (LC) in positive mode. This combination allows us to cover various DOM compounds ranging from highly cationic to highly anionic molecules. On-the-fly mass calibration of the Orbitrap at every scan by utilizing the “lock mass” function in the OT-FT-MS assures high mass accuracy throughout the whole sample run by a post-column introduction of internal labeled standards. A pooled quality control sample was used to increase reproducibility by applying systematic error removal using the random forest (SERRF) technique. The on-the-fly internal calibration application was tested on standard mixes before applying to actual DOM samples. In LC, the mass accuracy for the standards was consistently below +1.0ppm, and in IC below -4.0ppm, giving the OT-FT-MS the potential of reaching the massaccuracy of the Fourier-transform ion cyclotron resonance mass spectrometer (FT-ICR-MS). Once “lock mass” was established, estuarine surface water DOMsamples were analyzed in negative and positive modes. LC coupled to the OT-FT-MS for positive mode detection resulted in 915 compounds including 53 peptides and deaminated peptides that were identified through an in-house mass list comparison. IC coupled to the OT-FT-MS for negative mode detection resulted in 1432 extra compounds and including 361 deaminated peptides. Comparing the number of compounds detected shows the importance of analyzing both detection modes and the promising results of using IC in negative mode. A random forest-based normalization termed SERRF was compared to the traditional way of data normalization constant sum technique. Although constant sum provides already meaningful PCA data, SERRF assured a tight cluster of all Quality Control (QC). In IC mode, the relative standard deviation of QC samples was as high as 34.1% and corrected to 4.4% with SERRF, and in LC, the relative standard deviation was corrected from 28.1% to 3.9%. A combination of IC, LC, “lock mass”, and SERRF with the OT-FT-MS provides reliable and comprehensive approaches for DOM molecular characterization.