Characterization of lower Phong river dissolved organic matters and formations of unknown chlorine dioxide and chlorine disinfection by-products by Orbitrap mass spectrometry

Highlights

• Molecular changes in dissolved organic matter (DOM) along the river were tracked.

• Domestic discharges release DOM with high molecular weight (MW) and less oxidized.

• A natural lake contributed more oxidized DOM into the river.

• Water treatment further modified DOM to be lower MW.

• Unique and shared CHO and CHOCl disinfection byproducts (DBPs) were found by chlorine or chlorine dioxide.

Abstract

Dissolved organic matter (DOM) have been reported as precursors of disinfection byproducts (DBPs) and its molecular characteristics are rarely investigated due to its complexity. In this study, changes in the characteristics of DOM were investigated in the lower Phong River in Thailand in dry season and after the first rain in rainy season, using a non-targeted analysis with Orbitrap mass spectrometry. The river was rich with CHO features dominated by lignin-like molecules, while lipid-like molecules increase after domestic wastewater discharges. Wastewater discharge released DOM with higher molecular weight (MW) that was less oxygenated (low O/C) and less oxidized (low carbon oxidation state [C_os]). A lake affected by anthropogenic activities contributed more oxidized DOM into the river, while surface runoff carried DOM that is more oxygenated (high O/C), less hydrogenated (low H/C), and more oxidized (high C_os) to the stream. Water treatment further modified DOM to be lower MW. Approximately three hundred Cl-containing features (CHOCl) detected upstream were also found downstream. Disinfection by chlorine (Cl₂) or chlorine dioxide (ClO₂) formed both CHO and CHOCl DBPs. Low chlorine dosage applied to upstream and downstream samples resulted in many common unknown DBPs while increasing chlorine dosage resulted in more unique DBPs. At the same dosage, Cl₂ reacted with DOM more than ClO₂, including more oxidized molecules that are refractory to ClO₂. Both Cl₂ and ClO₂ produced chlorinated and non-chlorinated DBPs, and some DBPs were commonly found by both disinfections_. Cl₂-produced DBPs were more unsaturated (higher [DBE–O]/C) and oxidized (higher C_os) than ClO₂-DBPs.

Introduction

Natural dissolved organic matter (NOM) is a complex mixture of organic matter contained in natural water resources. As it is one of the main constituents in raw (untreated) water entering water treatment plants (WTPs), NOM characteristics have been widely studied using numerous methods and tools (Matilainen et al., 2011). NOM is harmless, but it is refractory and problematic to water treatment processes (Jacangelo et al., 1995). In addition, NOM especially dissolved NOM remains in conventional water treatment until water disinfection. In recent years, the emergence of synthetic organic compounds in aquatic environments has become a global issue and raised concerns about the aquatic environment. A wide variety of emerging organic pollutants (EOPs) especially pharmaceutically active compounds (PhACs) and endocrine-disturbing compounds (EDCs) have been reported in surface waters and groundwater at various concentrations ranging from nanogram to microgram per liter (Ternes, 1998; Kolpin et al., 2002; Nakada et al., 2007; Liu and Wong, 2013; Carmona et al., 2014). This pollution increases the complexity of dissolved organic matter (DOM) in natural water resources, and less than 10% of DOM have been characterized by traditional methods (Sleighter and Hatcher, 2008). Additionally, the reactions between chlorine and EOPs produces chlorinated pollutants of increasing complexity and potential toxicity (Bedner and MacCrehan, 2006; Kamel et al., 2009; Kosaka et al., 2016; Kamata et al., 2017).

Water disinfection is one of the main treatment processes in all water treatment plants. Chlorine has been used globally in WTPs as a water disinfectant for several decades due to its capabilities for disinfection (Akin et al., 1982) and oxidation (taste and odor control) at low construction and operational costs. In addition, chlorine can remain in distribution systems to the point of use to maintain disinfection efficiency in the pipeline. However, reactions between NOM and chlorine form halogenated compounds, disinfection byproducts (DBPs), several of which are regulated as toxic substances (Richardson et al., 2007; Deborde and von Gunten, 2008). Even if several disinfectants were proposed and applied as alternatives to chlorine, those disinfectants still generate specific DBPs, and the health impacts of a large portion of those DBPs are still unknown (Amy et al., 2000). Several types of common DBPs such as trihalomethanes (THMs), haloacetic acids (HAAs), and haloacetronitriles have been proven genotoxic or carcinogenic (Richardson et al., 2007) and have been regulated in tap water quality standards worldwide (WHO, 2008). With awareness on the health effects of DBPs, alternative disinfectants have been widely used, such as chloramines, ozone, and chlorine dioxide, as they produce less of those toxic DBPs (Amy et al., 2000; Zhang et al., 2000; Hua and Reckhow, 2007; Han and Zhang, 2018; Gilca et al., 2020). The byproducts of chlorine dioxide (ClO₂) include chlorite and chlorate, which have very low THMs and HAAs compared to other disinfectants (Amy et al., 2000). However, the formations of organic halogens by use of chlorine dioxide with organic compounds have been reported (Zhang et al., 2000; Han et al., 2017) and reviewed (Gilca et al., 2020). Furthermore, Hua and Reckhow (2007) reported that over 80% of chlorine dioxide halogenated DBPs are still unknown, and those unknowns would be highly polar DBPs (Richardson and Postigo, 2011).

Unknown or non-target screening analysis by high resolution and accurate mass spectrometry (HRAM-MS) has been widely used to analyze complex DOM in aquatic environments (Sleighter and Hatcher, 2008; Sleighter et al., 2008; Gonsior et al., 2011; Urai et al., 2014; Lavonen et al., 2015; Pan et al., 2020; Zhang et al., 2020a, 2020b). With high resolution, several thousand or up to ten thousand DOM peaks can be obtained in a single analysis. Highly accurate mass to charge (m/z) DOM peaks could be sufficient to estimate chemical formulae, which could provide useful information in the understanding of DOM characteristics (Kim et al., 2003) such as hydrogen to carbon (H/C) and oxygen to carbon (O/C) ratios observed in a Van Krevelen diagram (Van Krevelen, 1950), degree of unsaturation ([DBE–O]/C) (Phungsai et al., 2018) and degree of oxidation (carbon oxidation state, C_os) (Kroll et al., 2011). This information provides details on the changes in DOM profile within water resources (Tremblay et al., 2007; Urai et al., 2014; Yuthawong et al., 2017) and is widely used in characterizing DOM changes that occur as a result of water treatments (Gonsior et al., 2014; Lavonen et al., 2015; Phungsai et al., 2018, 2019, 2021; Phatthalung et al., 2020; Rakruam et al., 2021). In addition, detection of unknown DBPs formed during water chlorination is possible by this unknown screening technique (Lavonen et al., 2013; Gonsior et al., 2014; Phungsai et al., 2018, 2019; Rakruam et al., 2021). Detection of unknown DBPs could reveal the fates of putative precursors by different treatment processes (Phungsai et al., 2016, 2018).

This research investigated changes in molecular DOM from water sources to the tap by unknown screening analysis using Orbitrap mass spectrometry. Water samples were collected from the lower Phong River, Thailand, widely utilized for industrial, agriculture, fisheries, and drinking water and receives multiple treated and untreated discharges. DOM compositions and molecular characteristics along the river were investigated during summer and after the first rain. Unknown DBP formation by chlorination and chlorine dioxide disinfection of upstream and downstream water samples were analyzed, and changes in molecular characteristics were investigated in a semiquantitative manner.