Frequency-based paired mass distance method revealed the transformation pathway selection of organic compounds during mineralization treatment.

An in-depth understanding of organic compound transformations during wastewater treatment is crucial for guiding treatment upgrades. However, identifying key reactions in complex wastewater precisely and comprehensively remains challenging. Here, we developed a frequency-based paired mass distance (F-PMD) method to uncover these transformations, validated it in treatment of real waste brine, and used interpretable machine learning (IML) to reveal the effects of molecular properties. F-PMD identifies key reactions from Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) data by extracting high-frequency mass distances. The method was applied to a pH self-regulated ultraviolet-chlorine (UV/Cl2) system, where the pH self-regulation strategy increased total organic carbon (TOC) removal from 56.47 % to 82.13 %. The F-PMD results captured the most predominant reactions, covering > 70 % of precursors and products, far exceeding the coverage of conventional PMD results (minimum 46.72 %). The selection between carbon-loss and carbon-addition reactions was revealed as the core determinant of mineralization performance. High selectivity for carbon-loss pathways, which was promoted by high ·OH concentration and by compounds with a higher carbon (C) number, lower nominal oxidation state of carbon (NOSC), nitrogen to carbon ratio (N/C), chlorine to carbon ratio (Cl/C), and modified aromaticity index (AI_mod), enhanced mineralization. Collectively, these findings provide an advanced method for identifying key organic transformations in complex wastewater and offer mechanistic guidance for improving mineralization.