Formula Assignment Algorithm for Deuterium-Labeled Ultrahigh-Resolution Mass Spectrometry: Implications of the Formation Mechanism of Halogenated Disinfection Byproducts

Abstract

The ultrahigh-resolution mass spectrometry (UHR-MS) coupled with isotope labeling is of increasing attentions in elucidating the transform mechanisms of dissolved organic matter (DOM).However, there is a paucity of automated formula assignment algorithm applicable to halogenated disinfection byproducts (Xn-DBPs), particurally for iodinated organic compounds, and deuterated DOM . Herein, for the first time, we have developed a novel formula assignment algorithm based on deuterium-labeled UHR-MS, namely FTMSDeu, and the algorithm was applied to determine precursor molecules of Xn-DBPs and evaluate the relative contribution of electrophilic addition and electrophilic substitution reactions in Xn-DBPs formation according to the hydrogen/deuterium exchange of DOM molecules. Furthermore, tandem mass spectrometry with homologous-based network analysis was employed to validate the formula assignment accuracy (41%) of FTMSDeu for iodinated disinfection byproducts (In-DBPs). And the remaining In-DBPs compounds were assigned with the empirical rule of minimum number of non-oxygen heteratoms.The electrophilic substitution accounted for 82%-98%, 71%-89%, and 43%-45% of Xn-DBPs formation for Xn-DBPs containing chlorine, bromine, and iodine, respectively, manifesting the dominant role of electrophilic substitution in chlorine disinfection under conditions of low bromine and iodine concentrations. The absence of presumed Xn-DBPs precursors in some treatments in this study also suggests that Xn-DBPs formation include secondary reactions (e.g., oxidation, hydrolysis) in addition to electrophilic addition and/or substitution of halogens. These findings highlight the significance of isotopically labeled UHR-MS techniques in revealing the transformation of DOM in natural and engineered systems.

Development of an Internal Calibration Algorithm for Ultrahigh-Resolution Mass Spectra of Dissolved Organic Matter

“…The chemical formulas of all ESI(−)-FTICR-MS spectra (internally calibrated with DataAnalysis and FTMSCalibrate) were assigned by the FTMSDeu Code with the following calculation parameters: (1) signal-to-noise ratio ( S / N ) ≥ 6 and 10 for nonhalogenated and halogenated formulas, respectively; (2) an integer double bond equivalent (DBE) (≥0); 0.3 ≤ (H+Cl+Br)/C ≤ 2.25 and 0 < O/C ≤ 1.2 for formulas with C≥ 5; 0.3 ≤ (H+Cl+Br)/C ≤ 4 and 0 ≤ O/C ≤ 1.2 for formulas with 1 ≤ C ≤ 4; 1 ≤ 12 C ≤ 50, and 13 C ≤ 3; 18 O ≤ 1; −10 ≤ DBE-O ≤ 10; 14 N ≤ 5; 15 N ≤ 1; 32 S ≤ 3; 33 S ≤ 1; 34 S ≤ 1; P ≤ 1; 35 Cl ≤ 5; 37 Cl ≤ 5; 79 Br ≤ 5; 81 Br ≤ 5. For the ESI­(+)-FTICR-MS spectra (internally calibrated with DataAnalysis and FTMSCalibrate), the corresponding chemical formulas were assigned by the FTMSDeu Code with an additional function capable of identifying Na/K adducts.…”

“…The calculation parameters were the same as those for the ESI(−)-FTICR-MS spectra, except for the formulas with 0.3 ≤ (H+Cl+Br)/C ≤ 2.25 and 0 ≤ O/C ≤ 1.2. Molecular classes were categorized in the van Krevelen diagram according to the reported criteria …”

“…Data Analysis. The chemical formulas of all ESI(−)-FTICR-MS spectra (internally calibrated with DataAnalysis and FTMSCalibrate) were assigned by the FTMSDeu Code with the following calculation parameters: 10 For the ESI(+)-FTICR-MS spectra (internally calibrated with DataAnalysis and FTMSCalibrate), the corresponding chemical formulas were assigned by the FTMSDeu Code with an additional function capable of identifying Na/K adducts. All the adduct formulas with 13 C-isotopic peaks were validated using the 13 C-isotopic pattern.…”

“…Molecular classes were categorized in the van Krevelen diagram according to the reported criteria. 10 We evaluated the performance of the FTMSCalibrate Code in terms of the reproducibility of the chemical formula assignment by comparing it with a benchmark calibration method using DataAnalysis and the true positive ratio (TPR; calculated by eqs S1 and S2). 9 For statistical analysis, significantly different groups were determined using the oneway analysis of variance (ANOVA) when the parameters of interest passed the Kolmogorov−Smirnov normality test.…”

“…Generally, aquatic and terrestrial DOM samples contain recalcitrant compounds, such as refractory carboxyl-rich alicyclic molecules , and naturally occurring fatty acids . Therefore, the FTICR-MS spectra of DOM samples can be internally calibrated using a well-defined DOM homologous series via peak-by-peak comparison using the DataAnalysis software. ,− The manual internal calibration of tens and hundreds of FTICR-MS spectra using DataAnalysis is time-consuming. Therefore, the automated internal calibration using other methods is being investigated for the methodological study of DOM using FTICR-MS …”