Temporal Dynamics and Intermediate Product Formation in DOM Phototransformation Revealed by Liquid Chromatography Ultrahigh-Resolution Mass Spectrometry

Abstract

The complex composition of dissolved organic matter (DOM) has been extensively studied by modern high-resolution analytical methods. However, DOM reactivity is still enigmatic due to a lack of experimental data with sufficiently high temporal resolution to resolve the intrinsic dynamics within DOM. Likewise, extensive isomeric overlap prevents studying transformation of DOM components with respect to their chemical properties, e.g., molecular polarity. Online ultrahigh-performance liquid chromatography with ultrahigh-resolution mass spectrometry (UHPLC-UHRMS) increases the resolution of isomeric DOM composition across a wide range of polarity. We performed a TiO₂-aided photo-irradiation experiment with wastewater treatment plant effluent with high temporal sampling resolution (8 time points, 5 h irradiation). Besides new products (<10%) and removed components (25-60%), intermediate products (IntP) were also found, representing 20-60% of components within distinct polarity fractions. The reaction time to reach the peak magnitude maximum was positively related to the H/C ratio of IntP. About 35% of the DOM components showed different reactivities for different polarity fractions. If applied to experiments in the future, our approach offers new perspectives for biogeochemical interpretation and provides important information for drinking water processing or wastewater treatment with respect to potential toxic IntP.

Keywords: LC-FT-ICR-MS; WWTP effluent; biogeochemical cycling; dissolved organic matter; high time resolution; intermediate products; photo degradation.

1

Flowchart for the experiment, mass spectrometry measurement and data evaluation steps.

2

(A) Examples for reaction time courses for a molecular formula with reactivity classes produced (Prod, green), degraded (Degr, red), intermediate product (IntP, blue), and resistant (Res, black). δRAW = 0.265 = 26.5% is shown as the minimum threshold for significance. (B) Example for an IntP (with maximum RAW peak magnitude at t₄) and designation of peak magnitudes at reaction time points.

3

(A) Distribution of reactivity classes (as fraction of all molecular formula, MF) as a function of retention time (RT). Reactivity classes were combined as described in STEP 4. (B–D) Distribution of MF classes (CHO, CHNO, CHOS, CHNOS) of assigned products (B: Prod), degraded MF (C: Degr) and intermediate products (D: IntP). Note that segment at 20.5 min was excluded from further discussion due to the high number of MF classified as r.n.a. The corresponding data for simulated pseudo-DI-FT-ICR-MS data are shown in the orange box, as described in Section .

4

Reaction time courses of specific effluent DOM molecular formulas (MF) revealed by LC-FT-ICR-MS. (A) C₇H₁₀O₅, classified as Prod (e.g., RT = 11.5 min), (B) C₁₁H₁₂O₇S₁, classified as Degr (e.g., RT = 11.5 min), (C) C₈H₈O₄S₁, classified as IntP (e.g., RT = 12.5 min), and (D) C₁₂H₁₄O₈S₁, with opposite reactivity at RT = 10.5 min (Degr) and at RT = 12.5 min (IntP).

5

Distribution of assigned reactivity classes at RT = 10.5 min based on the molecular H/C vs O/C ratios. (A) CHO and (B) CHNO. Distribution of IntP as a function of the (reaction) time point where the maximum peak magnitude was observed for (C) CHO and (D) CHNO. The distributions of all reactivity classes of all RTs are shown in Section SI7.

6

Comparison of reactivity class distribution, DI 8TP model vs LC model. (A) Distribution of reactivity classes in the LC model (Prod, green; Degr, red; IntP, blue; Res, gray; r.n.a., brown) for molecular formulas (MFs), which have been assigned as products in the DI 8TP model (“DI Prod), (B) distribution of degraded MF (”DI Degr”), (C) intermediate products (“DI IntP”), and (D) corresponding distribution of resistant MF (“DI Res”). Inset pie charts indicate the distribution of reactivity classes in the DI 8TP data set and values on bars indicate the number of MF of the respective DI 8TP reactivity class detected in each LC segment. Cf. data provided in es5c01986_si_002.xlsx, Sheet SI x.3.