1,4-Dioxane contamination of German drinking water obtained by managed aquifer recharge systems: Distribution and main influencing factors

Journal Pre-proofs 1,4-Dioxane Contamination Of German Drinking Water Obtained By Managed Aquifer Recharge Systems: Distribution And Main Influencing Factors Ursula Karges, Diana Ott, Sabrina De Boer, Wilhelm Püttmann PII: DOI: Reference:

S0048-9697(19)34774-6 https://doi.org/10.1016/j.scitotenv.2019.134783 STOTEN 134783

To appear in:

Science of the Total Environment

Received Date: Revised Date: Accepted Date:

28 July 2019 30 September 2019 1 October 2019

Please cite this article as: U. Karges, D. Ott, S. De Boer, W. Püttmann, 1,4-Dioxane Contamination Of German Drinking Water Obtained By Managed Aquifer Recharge Systems: Distribution And Main Influencing Factors, Science of the Total Environment (2019), doi: https://doi.org/10.1016/j.scitotenv.2019.134783

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Elsevier B.V. All rights reserved.

1,4-DIOXANE CONTAMINATION OF GERMAN DRINKING WATER OBTAINED

BY

MANAGED

AQUIFER

RECHARGE

SYSTEMS:

DISTRIBUTION AND MAIN INFLUENCING FACTORS

Ursula Karges*, 1, Diana Ott1, Sabrina De Boer1,2, Wilhelm Püttmann1 1Institute

of Atmospheric and Environmental Sciences, Goethe–University Frankfurt am Main,

Altenhöferallee 1, 60438 Frankfurt am Main, Germany. 2Department

of Chemical Engineering, School of Engineering, Universidade de Santiago de

Compostela, Rúa Lope Gómez de Marzoa, s/n, 15782 Santiago de Compostela, Spain

*Correspondence

Ursula Karges Institute of Atmospheric and Environmental Sciences, Goethe-University Frankfurt am Main, Germany. Tel.: +49-(0)69-798-40232 E-Mail: [email protected]

ABSTRACT 1,4-Dioxane, a cyclic ether that has been classified as a class 2B carcinogen by the US-EPA, is a substance of growing environmental concern because of its abundant occurrence in surface waters worldwide. Its high polarity and low biodegradability hamper its retardation in aquifer systems. Previous investigations in Germany have shown that 1,4-dioxane is already widely distributed in rivers and can be found in groundwater at contamination sites. Therefore, the present study shall provide an overview of the Germany-wide distribution of 1,4-dioxane in finished drinking water (FDW) obtained by managed aquifer recharge (MAR) systems. Thus, we investigated the 1,4-Dioxane levels in FDW obtained by MAR, such as river bank filtration (RBF) or artificial groundwater recharge (AGR), in regions that are supplied by surface water bodies (mainly rivers) with already known 1,4-dioxane contaminations. In total, 125 FDW samples and 33 samples of corresponding surface waters were analyzed for 1,4-dioxane content using solid phase extraction followed by gas chromatography–mass spectrometry (SIM-mode) using a slight modification to US-EPA method 522. About 80% of the investigated FDW samples contained 1,4-dioxane at levels exceeding the limit of quantification (0.034 μg/L); the maximum value was 2.05 μg/L. However, a maximum concentration of 3 μg/L was obtained in the surface water samples. Three main factors were associated with elevated levels of 1,4-dioxane in the FDW: A significant 1,4-dioxane contamination of the associated surface water, the application of RBF instead of AGR, and the proportion of available unpolluted groundwater and/or reservoir water blended in the individual waterworks. The results show that 1,4-dioxane should be critically monitored during FDW production by means of MAR not only in Germany. The

findings are also of relevance to neighboring countries depending on the same river systems and for research in the field of small mobile substances in drinking water production in general. KEYWORDS: 1,4-Dioxane, Drinking water, Bank filtration, Artificial groundwater recharge, Germany

1 INTRODUCTION Pollution of drinking water by 1,4-dioxane was first identified in the 1970s, and 1,4-Dioxane is now regarded as one of the "emerging pollutants": a group of substances whose occurrence in the environment has been the subject of intensive research since the late 1990s (Adamson et al., 2017; Carrera et al., 2019; Halden, 2015; Mohr et al., 2010; Sauvé and Desrosiers, 2014). Various toxicity studies have given reasons for assuming that 1,4-Dioxane can negatively affect human health. Consequently, the substance has been classified by the US-Environmental Protection Agency (US-EPA) as likely to be carcinogenic to humans (class 2B) (Mohr et al., 2010; US-EPA, 2013). Knowledge about its fate and behavior in the aquatic environment is therefore of considerable interest for human exposure assessment and water protection control in general (Adamson et al., 2017; EC, 2002; Mohr et al., 2010). 1,4-Dioxane has primarily been applied as a stabilizer for chlorinated solvents such as 1,1,1trichloroethane (TCA). The Montreal Protocol banned the use of TCA from 1996 because it is one of the compounds responsible for ozone depletion in the stratosphere (Mohr et al., 2010; UNEP, 2006). However, 1,4-dioxane is still present as a co-contaminant in groundwater plumes of TCA (Mohr et al., 2010; U. S. EPA, 2016). Previous studies have reported a correlation between the occurrence of 1,4-dioxane and the chlorinated solvent trichloroethylene (TCE) in groundwater systems in the USA and in Germany. This indicates that 1,4-dioxane was not only used as stabilizer for TCA, but also for TCE (Anderson et al., 2012; Karges et al., 2018). In addition, 1,4-dioxane is a commonly used industrial solvent in, for example, paints, plasticizers, and coatings (Mohr et al., 2010). The substance is also generated as an unintended by-product from the production of personal care products and detergents during the ethoxylation of alcohols and the sulfation of alcohol ethoxylates to ether sulfates (Ortega, 2012), and it is also

formed as an unintended by-product during the esterification reaction in polyester production (Zenker et al., 2003). The multipurpose, ongoing usage of 1,4-dioxane, and it diverse means of (un-intended) formation, results in a multitude of possible entry pathways into surface water and groundwater (Mohr et al., 2010). Previous studies demonstrated that effluents from waste water treatment plants are among the most probable point sources for 1,4-dioxane in rivers (Abe, 1999; Antoniou et al., 2006; Glassmeyer et al., 2005; Stepien et al., 2014). Because of its biological recalcitrance, its miscibility with water, its low evaporation from the water phase (Henry`s law constant: 4.80 X 10-6), its polarity and its low tendency for absorption to sediments and biota (Log KOW: 0.27) (US-EPA, 2013), 1,4-dioxane is an extremely challenging substance to remove in water treatment (Adamson et al., 2017; Mohr et al., 2010; US-EPA, 2013; Triplett Kingston et al., 2010; Vescovi et al., 2010; Zenker et al., 2003; Zhang et al., 2017). Conventional, mainly physical water treatment processes (such as sand filtration, flocculation, precipitation and sedimentation, as well as adsorption by activated carbon) have been proven to be ineffective for the sufficient removal of 1,4-dioxane from water (DiGuiseppi et al., 2016; Mohr et al., 2010; Zenker et al., 2003). A sufficient removal of 1,4-dioxane from water has been proven for advanced oxidation processes (AOP), such as ozone/hydrogen peroxide (O3/H2O2), UV/H2O2, O3/H2O2/UV, TiO2/UVA, Fe/H2O2 (Fenton`s reagent) and Fe(0)-persulfate (Adams et al., 1994; Barndõk et al., 2014; Carrera et al., 2019; Jasmann et al., 2016; Kambhu et al., 2017; Mohr et al., 2010; Suh and Mohseni, 2004, 2004). Synthetic media, such as AMBERSORB™ 560, have been shown to be effective in removing varying concentrations of 1,4-dioxane from aqueous waste streams (Han et al., 2012; Jasmann et al., 2016; Woodard et al.,

2014). However, the application of these methods is often significantly more costly than traditional physical treatment methods. Because of the negative health effects of 1,4-dioxane and its persistence in the aqueous environment, several countries have implemented initial monitoring programs and adopted screening levels for 1,4-dioxane in water. The World Health Organization has established a guideline value of 50 μg/L for 1,4-dioxane in potable water (WHO, 2011). This guideline has been adopted by Japan (Wakayama, 2003; WHO, 2011). Other countries have launched more stringent guideline values. The US-EPA, for example, calculated a tap water screening level of 0.67 µg/L, based on a 1 × 10−6 cancer risk level, and established a drinking water advisory level of 0.35 μg/L (US EPA, 2018; US-EPA, 2013; US-EPA IRIS, 2010). The values set as a limit for drinking water by individual US states vary by three orders of magnitude, but tend to be stricter than the US-EPA level (Suthersan et al., 2016). In contrast, in many countries, a 1,4-dioxane standard or guideline value for drinking water and groundwater is still entirely lacking. In Germany, 1,4-dioxane has not been regulated within the drinking water ordinances, but a guideline value of 5 µg/L is recommended by the German Federal Environment Agency (UBA) (LFU Bayern 2016). In general, German finished drinking water (FDW) quality is particularly high, because of the abundant availability of groundwater. About 70% of the German FDW is supplied by these reserves, whereas the share of water originating from managed aquifer recharge (MAR) systems is about 17% (Supporting Materials Table S1). So far, FDW in Germany has only occasionally been examined for 1,4-dioxane on a district basis. The results show that 1,4-dioxane is a pollutant of groundwater and river waters, and also of some bank filtration sites (Karges et al., 2018; LfU Bayern, 2016; Stepien et al., 2014, 2013).

Considering the incomplete picture of 1,4-dioxane pollution in German FDW when compared with other countries (Adamson et al., 2017; Hamann et al., 2016a), a comprehensive examination is urgently required, along with the identification of the key contamination factors. It is especially important to evaluate the connection between the usage of bank filtrate for FDW supply and elevated 1,4-dioxane levels in FDW (LfU Bayern, 2016). Detailed knowledge of the distribution and sources of 1,4-dioxane contamination in the German FDW, including the contamination of relevant surface waters, is of relevance to the health of the German population, and also that of neighboring countries. For example, a part of the dutch drinking water is obtained from MAR water from the Rhine in the Netherlands, whereas the Austrian FDW demand is partly covered by MAR water from the Danube (Hamann et al., 2016a; Handl et al., 2017; Stuyfzand, 2016; Weigelhofer et al., 2013). In general, the behavior of water-mobile 1,4-dioxane is representative of the behavior of a variety of other, emerging organic micropollutants, and is therefore of interest for research into drinking water quality in general (Adamson et al., 2017; Farré et al., 2008; McElroy et al., 2019; Sauvé and Desrosiers, 2014; Schriks et al., 2010; Stepien et al., 2013). In the present study, 1,4-dioxane was analyzed in FDW throughout Germany, and was collected preferentially in regions where FDW is obtained from MAR.

2 INVESTIGATION AREAS AND METHODS 2.1 Regions of interest A study of drinking water quality in Germany has to take into account the fragmented nature of the state of drinking water supply in Germany, which has traditionally been the responsibility of individual cities and municipalities. Drinking water supply is therefore often very heterogeneous and determined by historical factors. A variety of factors are taken into account when decisions are made regarding what percentage of MAR water a particular FDW contains and which MAR system is preferred. The investigated cities, municipalities, and surface water bodies are located in the federal states of Berlin, Saxony, North Rhine-Westphalia, Rhineland-Palatinate, Hesse, Baden-Wurttemberg, and Bavaria (Figure 2 and Supplementary Materials Table S4, S5). The spatial origin of the samples is shown in Figure 2.

2.1.1 Geographical setting and regional conditions Because of generally abundant groundwater resources, the public water supply in Germany is largely based on groundwater. Therefore, the quantitative distribution of groundwater resources, as well as their accessibility for water extraction, are key factors for FDW supply (BMU, 2008). Large parts of Germany have sufficient groundwater resources to cover the existing drinking water requirements (BMU, 2008; UBA, 2010). However, in some areas, for example in North Rhine-Westphalia or Saxony, the groundwater level has dropped considerably as a result of centuries of mining activity (UBA, 2016). In the city of Dusseldorf in North Rhine-Westphalia, for example, meeting water demand by means of river bank filtrate (RBF), has been a longstanding tradition (Umweltamt Düsseldorf, 2017). The subsoil here is composed of loosely

layered Quaternary sands and coarse gravel of the lower Rhine-terraces. The high permeability of the soil layers offers good conditions for the continuous extraction of raw water via well galleries. However, it also leads to a higher susceptibility to mobile and hydrophilic pollutants because they cannot be sufficiently retained during the rapid passage through the soil (Umweltamt Düsseldorf, 2017). In Saxony, most of the drinking water demand is covered by lake and reservoir water, but at least 26% is met by MAR water from different sources. The respective share of water from MAR varies greatly, even between individual districts of Leipzig and Dresden, for instance. Because of their high population density, urban agglomerations have a high water requirement that cannot always be met by existing groundwater resources, such as in parts of Bavarian Franconia, the Rhine Main area, and Berlin (Krause, 2014; Massmann et al., 2007; WVR, 2017; Ziegler, 2001). In Berlin, the water requirement of the metropolis is almost completely ensured by MAR (Schulze, 1997; Ziegler, 2001). Water cycles here are often narrow, partially closed systems of surface water and drinking water sources (Jekel et al., 2013). The metropolitan area around Nuremberg is also an agglomeration for which the water supply can not only be ensured via the local groundwater resources. It is located in an area with a low annual precipitation; the subsoil consists predominantly of hard rock such as granite, sandstone, or limestone, which are unfavorable for the recharge of the groundwater. The water supply to this area is met by means of water pipelines that transport RBF water from the Lech estuary to Franconia (VEBW, 2013). The RBF is obtained here from wells of up to 15 m in depth, located in Quaternary sediments (Schielein and Schellmann, 2016). In the Rhine Main area, the high water demand of the Hessian Ried region is added to the large demand for water from the metropolitan area around Frankfurt am Main. Hessian Ried, for example, has been an important agricultural area for centuries, and the demand for water has been correspondingly high. To protect groundwater levels, the groundwater in this region is

supplemented with pre-purified Rhine water (LDEW, 2013; WVR, 2017). Hessian Ried belongs geologically to the northern Upper Rhine Graben, and has Quaternary and Tertiary deposits; the permeable sandy to gritty sandy layers are of particular importance in terms of water management. In this area, AGR is used for FDW supply (WHR, 2018, 2017).

2.2 MAR Systems Two different MAR systems are applied in Germany: river bank filtration (RBF) and artificial groundwater recharge (AGR) (Figure 1). In both cases, the water is composed partly of groundwater and partly of filtrated surface water. These proportions can vary significantly in different locations because of differences in hydrological conditions.

Figure 1. MAR systems: simplified representation of the principal differences in water production from bank filtration (RBF) and (B) artificial groundwater recharge (AGR)

During RBF, most pollutants present in the river water are eliminated by a variety of processes. The pollutants are subject to sedimentation and filtration, chemical precipitation, and sorption and ion exchange, as well as microbial degradation (Zenker et al., 2003). Two main factors significantly influence the amount of recoverable bank filtrate: the infiltration area adjacent to the river, and the permeability of the aquifer (Polomčić et al., 2013; Schmidt et al., 2003). RBF is often used as a first major treatment process for surface water for subsequent FDW production (Hiscock and Grischek, 2002). Ozonation and active charcoal filtration are often applied as follow-up treatment steps (Achten et al., 2002). The decision on the implementation of additional treatment steps strongly depends on the type and amount of pollution in the raw water (Hamann et al., 2016). RBF followed by ozonation and active charcoal filtration has been documented to be very effective in the removal of most contaminants present in river water. However, according to results from recent studies (Barndõk et al., 2014; Stepien et al., 2014), the efficiency of this combination of processes for 1,4-dioxane is questionable. High levels of river contamination may pose a risk to FDW quality obtained by RBF close to a river (Hiscock and Grischek, 2002). Sorption, biodegradation, and distance/residence time are the main factors that determine the efficiency of the removal of biodegradable and adsorbable organic water constituents during subsurface transport (Hamann et al., 2016; Hiscock and Grischek, 2002; Schriks et al., 2010). However, RBF has been shown to be ineffective for the removal of a small number of substances, such as carbamazepine and primidone, and in particular those of a strong hydrophilic behavior and small molecular size (Rauch-Williams et al., 2010). In an comprehensive study of 247 compounds, 1,4-dioxane and nine other substances showed fully persistent behavior even after 3.6 years of transit time (Hamann et al., 2016). In particular, ether compounds such as methyltert-butylether and 1,4-dioxane may withstand even the ozonation step and are therefore among

the most challenging compounds in terms of FDW production from river water using RBF (Achten et al., 2002; Stepien et al., 2014). AGR is applied at locations where the FDW demands are not covered by groundwater and/or RBF. The process of AGR involves irrigation, or infiltration basins, or channels (Figure 1). This method allows for the control of the amount of infiltration and allows pre-treatment of the surface water to standards suitable for infiltration (Polomčić et al., 2013; Schmidt et al., 2003).

2.3 Sampling The samples analyzed in this study were collected during several campaigns in the period from April 2015 to February 2016. In a pre-study, FDW samples from cities that are not supplied by MAR systems were revealed to contain 1,4-dioxane at levels exclusively below the LOD. Therefore, FDW samples were collected only in areas in Germany where surface waters treated by MAR systems are a significant source of FDW. Depending on accessibility, FDW samples were taken from taps in buildings after a minute pre-run, collected directly in prepared glass bottles, and stored in the dark at 4 °C. In addition, surface water samples were collected from corresponding source waters. Wherever possible, these samples were collected from bridges by means of a metal sampling vessel. If this was not possible, samples were collected from the shore of the water body. In both cases, the water samples were immediately transferred to the same type of prepared bottles as the drinking water samples. To get an overview, the main study areas and sampling sites are shown on the map (Fig. 2).

Figure 2. Locations of the main study areas for the FDW and surface water sampling. Please note: only investigated cities, rivers, and states that contain an investigation area are named.

Depending on accessibility, the number of samples studied varied in each area and city. The respective numbers are shown in the table (Tab. 1).

Table 1 Number of samples (n) taken per region

City/Region

n

Berlin

12

Dresden

8

Leipzig area

12

Ruhr-area

13

Cologne

2

Düsseldorf

9

Mainz&Rhine-Hesse

8

Franken

24

Passau

4

Rhine-Main-area

27

Baden Wurtemberg

5

Mecklenburg-Western Pomerania

1

2.4 Experimental

2.4.1 Chemicals

1,4-dioxane (99.9%) and 1,4-dioxane-d8 (99%) were obtained from Sigma-Aldrich (Steinheim, Germany). 4-chlorotetrahydropyrane (96%) was purchased from TCI (Tokyo, Japan). 4chlorotetrahydropyrane was used as an internal standard (IS) and 1,4-dioxane-d8 as a surrogate

(SU). Analytical grade dichloromethane (distilled before use) and hyper-grade methanol were both obtained from Merck (Darmstadt, Germany). An Astacus ultrapure water purification system, from MembraPure (Bodenheim, Germany), equipped with a TwinPak cartridge (with organic scavenger) from Purefekt (Karlsruhe, Germany), was used to produce ultrapure water. Stock solutions of 1,4-dioxane and of the SU were prepared in methanol, each at a 1 μg/µL concentration. The IS stock solution was prepared in dichloromethane, also at a 1 μg/µL concentration. Working standard solutions and calibration curves were prepared using appropriate dilutions of stock solutions in methanol or dichloromethane. Sodium bisulfate and sodium sulfite (98%) were purchased from Sigma Aldrich Chemistry (Steinheim, Germany). Anhydrous sodium sulfate was supplied by Sigma Aldrich (Seelze, Germany) and baked out at 400 °C for 4 h before use.

2.4.2 Extraction and Analysis

A slightly modified version of the US EPA method 522 for the determination of volatile organic compounds in drinking water was used in this study for extraction and analytical method (Karges et al. 2018; Stepien and Püttmann, 2013; Stepien et al., 2014). The method only requires water samples of 100 ml and is based on solid phase extraction (SPE) using “Supelclean™ ENVICarb™ Plus” SPE cartridges (Supelco, bed wt. 400 mg, 1 mL) followed by gas chromatography– mass spectrometry (GC-MS) analysis. For GC-MS analysis, a Trace 2000 Ultra GC system coupled to a Voyager quadrupole MS instrument (EI, electron ionization) (ThermoQuest Finnigan, Dreieich, Germany) was operated in selected ion monitoring (SIM) mode with electron impact (EI) ionization at 70 eV. M/z = 88 (1,4-dioxane), m/z = 96 (SU; 1,4-dioxane-d8), and m/z = 55 (IS; 4-chlorotetrahydropyran) were monitored ions. The GC was equipped with a DB 624-

column (30 m×0.25 mm ID, 1.40 µm film thickness) (Agilent, Waldbronn, Germany). Using a Combi PAL autosampler (CTC Analytics, Switzerland), 2 µl of each 500 µl sample were transferred to the injector, which was operated at 220 °C in a splitless mode for 1 min followed by a split mode and a split flow of 1 mL/min. Helium was used as a carrier gas (purity >99.99%). All relevant parameters are summarized in 2 (Karges et al., 2018; Stepien et al., 2013). XCalibur software (Thermo Fisher Scientific, version 2.0.7) was used in order to process and calculate relevant data.

Table 2 GC setup Column

DB-624, 30 m×0.25 mm ID, 1.40 µm film thickness

Injector

220°C, 1 min splitless, He 1 mL/min

Analytical method

37°C -2.5 min, 4°C/min to 75°C, 10°C/min to 220°C -10 min

Source

220°C

Interface

250°C

Emission current

150 µm

Detector voltage

500 V

OriginPro 2018G software and MATLAB2018b software were used to perform the statistical analysis.

2.4.3 Quality assurance

Amber glass bottles used for sample collection were pre-cleaned with distilled water and methanol and subsequently heated in an oven at 110 °C for a minimum of 2 h. To prevent the

analytes from degradation by chlorine in the potable water samples, sodium sulphite was added to each bottle at 50 mg/L. Immediately after the sampling, a microbial inhibitor (sodium bisulfate) was added to each bottle at a concentration of 1 g/L, which is also in accordance with the requirements of US-EPA method 522 (US-EPA, 2008). Samples were stored in the dark in a refrigerated storage room at 4 °C for a maximum of seven days, pending further processing. Shortly before extraction, each of the 100 mL water samples and quality control standards were enriched with 5 μL of 1,4-dioxane-d8 (SU) to reach the final concentration of 10.0 μg/L in the extract. Control standards were spiked at 1 µg/L. In each extraction batch of twelve samples a method blank and a control standard sample (spike) were included, which were treated in the same way as the samples. Supelclean™ ENVI-Carb™ Plus cartridges were used for extraction. The acceptable recovery for the SU and spike was set between 70% and 130% and only such samples with a recovery of SU within these limits were further considered in the evaluation. The calibration curve ranged from 0.04 to 2.0 µg/L. The limits of quantification (LOQ) were 0.032 µg/L for ultrapure water, 0.034 µ/L for drinking water samples, and 0.048 for surface water samples from Lake Constance. The limit of detection (LOD) for drinking water and for ultrapure water was determined to be 0.011 µg/L, and for surface water was 0.016 µg/L. Blank values were below the limit of detection in each case. To verify the precision and accuracy of the method (SPE followed by GC/MS-SIM), a total of seven 100 mL ultrapure water samples spiked with 1,4-dioxane at 1 μg/mL were extracted and analyzed to determine the percentage of recovery and the percentage of relative standard deviation (% RSD). A mean recovery of 97.5% and an RSD of 4.3 % were determined.

3 RESULTS AND DISCUSSION

A total of 125 FDW samples and 33 samples taken from corresponding surface water bodies (rivers, lakes, and reservoirs) were analyzed for 1,4-dioxane content. The compound was identified in 98 out of the 125 FDW samples and in 27 out of the 33 surface water samples. The results of the FDW survey indicate strong regional differences in the levels of 1,4-dioxane concentrations across Germany (Figure 3 and Supplementary Materials Table S2).

Regional differences in FDW (and dependence on source waters)

Concentrations of 1,4-dioxane equal to or above the LOQ (0.034 µg/L) in the FDW samples ranged from low values of 0.034 µg/L in Darmstadt (Hesse) up to a maximum value of 2.05 µg/L in Erlangen (Bavaria). In addition to these regional differences, strong differences were also found within individual regions.

Figure 3. Concentrations of 1,4-dioxane (μg/L) in FDW and surface water bodies for all cities/areas investigated.

In the metropolitan area of Cologne/Düsseldorf, for example, the concentrations varied from below the LOD (Cologne) up to 0.72 µg/L (Düsseldorf) and the mean FDW levels of Düsseldorf and the Ruhr-area differ significantly at α=0.05 (unpaired t-test). Similar patterns were observed in the Rhine-Main area (
Low concentrations within the range from the LOQ (0.034 µg/L) to the LOD (0.01 µg/L) were predominantly determined in samples representing a blend of MAR water with groundwater and/or unpolluted reservoir water. Conversely, FDW samples with high concentrations of 1,4dioxane are associated with MAR systems located at water bodies that were revealed to be heavily contaminated with 1,4-dioxane. Therefore, the investigation of the water bodies associated with the analyzed FDW samples comprised a major part of this study. The influence of source water contaminated by 1,4-Dioxane on FDW quality becomes particularly evident when comparing the situation in the adjacent federal states of BadenWurttemberg and Bavaria. No 1,4-dioxane was detected in the FDW samples from Baden-Wurttemberg (Figure 3). All of the tested locations receive their water from Lake Constance, which is one of the largest surface water reservoirs in Germany (Petri, 2006). The determined concentrations of 1,4-dioxane in the two lake samples were below the LOD, which coincides with the results for the FDW samples. In contrast, high concentrations of 1,4-dioxane, with mean values exceeding 1 µg/L, were detected in a majority of the FDW samples collected from Bavarian cities and communities. In all these cases, a large proportion of the FDW can be traced back to one common origin. The FDW is mainly, or partly, comprised of RBF pumped near the estuary of the river Lech into the river Danube (Figure 5), with the Lech river water comprising the majority of the RBF (Krause, 2014; LfU Bayern, 2016). Several industrials sites of polyester producing or processing are located along the river Lech and one of tributaries, the river Wertach. These industrial sites may possibly be point sources of 1,4-dioxane, directly through their own sewage treatment plants or indirectly by the discharge from municipal sewage treatment plants into the river Lech or its tributaries prior to the time of our sampling (Schilling and Arznet, 2018). In fact, since 2017 (i.e. after the

sampling campaigns of the present study), some of these point source locations in Bavaria were identified and have been retrofitted with AOP methods (Arznet and Schilling, 2018), for example at the polyester production site in Bobingen on the river Wertach, a tributary of the river Lech. According to recent reports by the responsible authorities, this action led to an impressive decline in the 1,4-dioxane levels in sewage treatment discharges (LfU Bayern, 2019; Schilling and Arznet, 2018). Small-scale AOP treatment methods at industrial locations might, therefore, be a solution for reducing the emission of 1,4-dioxane into surface waters. The variation of concentrations of 1,4-dioxane in FDW samples containing more than 90% of MAR is high, with values ranging from 0.04 µg/L (Darmstadt/Gelsenkirchen resp.) to 1.53 / 1.72 µg/L (Treuchtlingen/Erlangen). This range is not surprising given that the concentration levels are highly dependent on the concentrations in the river water from which the MAR water is obtained. The results of the investigated surface water samples that correspond to the FDW samples were also in a wide range: from below the LOD (e.g. the Danube upstream of the Lech estuary) through to low concentrations of 0.05 µg/L (Main at lower Franconia) and up to 2.8 µg/L (Lech). For example, there was a significant (α = 0.05) difference between the measured 1,4dioxane value of the FDW from Erlangen (1.72 µg/L) which was directly sourced from RBF from the highly contaminated river Lech, when compared with the 1,4-dioxane value for FDW from Essen (0.11 µg/L) or Volkach (0.11 µg/L). FDW for Essen is sourced from MAR from the moderately contaminated river Ruhr (0.28 ±0.1 µg/L), while FDW for Volkach is sourced from the lightly polluted river Main (0.11 µg/L), respectively. To identify the sources of the 1,4-dioxane in the FDW, the rivers and surface water bodies used for MAR were also analyzed for 1,4-dioxane, and the concentrations were compared with the values determined for the FDW (Figure 4, 4C). The residence time of the river water during bank

filtration could have been different at each location. Therefore, it was not possible to identify the time span between the infiltration of the water from the river into the aquifer and the arrival of the water in the FDW.

Figure 4. Mean concentrations of 1,4-dioxane in FDW obtained by MAR, and in related surface water bodies; A: FDW mainly obtained by RBF and AGR from the river Rhine; B: Mean concentrations in FDW in Bavaria, mainly obtained by RBF from the rivers Lech and Danube. In the cases of Rottendorf, Marksteft, and Sulzfeld, RBF from the river Lech was blended with RBF from the river Main; C: Concentrations of 1,4-dioxane in source waters and in the FDW associated with the source waters.

In most of the cases shown, high concentrations of 1,4-dioxane in river water led to increased levels of 1,4-dioxane in FDW obtained from the river. The highest concentration of 1,4-dioxane was measured in water from the river Lech (2.8 µg/L). This resulted in a FDW level of 2.05 µg/L for the city of Erlangen, which is supplied mainly by RBF from the river Lech. More than 90% of Passau‘s FDW is obtained by MAR from river water. However, contamination of Passau’s FDW cannot be attributed exclusively to the contamination of a single river because the RBF catchment area for Passau is located three kilometers downstream of the confluence of the river Inn and the Danube. Even though the river water in this area is predominantly made up of Inn river water, the quality of Passau’s FDW is ultimately influenced by both rivers, and possibly by the contamination of the Lech through a cascading effect. At the time of the FDW sampling in Passau, the corresponding river samples yielded 1,4-Dioxane concentrations of 0.38–0.54 µg/L (Danube) and 0.63 µg/L (Inn). In the converse case, low concentrations of 1,4-dioxane in the rivers are in most cases reflected in low levels of 1,4-dioxane in the associated FDW (Figure 4). This is clearly evident, for example, in the samples from the Ruhr area. The FDW in the cities of Schwerte, Dortmund, and Essen consists of over 90% of AGR water obtained from the river Ruhr. In March 2015, the mean concentration of 1,4-dioxane in the water from the river Ruhr was 0.28 µg/L (standard deviation: 0.1 µg/L), which represents a moderate to low value for river water, and in June 2015, FDW samples from these cities showed low levels of 1,4-dioxane ranging from 0.08 to 0.1 µg/L. Comparable results were obtained from the FDW in the city of Volkach, located on the river Main, which had a low 1,4-dioxane content of 0.11 µg/L. The FDW in Volkach is exclusively obtained by RBF from the river Main. A 1,4-dioxane concentration of 0.11 µg/L was determined for a sample of river water from the catchment area that was taken at the same time as the FDW sampling. This concentration is confirmed by other data collected throughout this river section.

All data collected in this river section show low concentrations of 1,4-dioxane, with a background level of 0.05–0.09 µg/L. At this location a depletion of 1,4-dioxane by RBF and the subsequent soil passage was not visible. Because the residence time during the subsurface passage is generally quite heterogeneous throughout Germany (Storck et al., 2012) and unknown in this case, 1,4-dioxane concentration peaks in the river water are not completely excludable. These uncertainties can only be eliminated by regular monitoring. Overall, the samples from the river Main in Lower Franconia showed an average concentration of 0.09 µg/L, with a standard deviation of 0.04 µg/L at the time of sampling in 2015. In the study of Stepien et al. (2014), a much higher median concentration of 0.53 µg/L was reported for the river Main. This can be explained by the fact that, in that study, the river Main was only sampled in the Rhine-Main area close to the confluence of the Main and the Rhine, where the contamination of the river with 1,4-dioxane was revealed to be much higher than in the upper parts of the river in Lower Franconia. The river Main in Lower Franconia is among the least contaminated rivers in the present study with respect to 1,4-dioxane. Nonetheless, some districts of the city of Würzburg and various communities from Lower Franconia, such as Rottendorf, Marksteft, and Sulzfeld, obtain additional supplies of RBF from the river Lech via a pipeline of more than 100 km in length (Figure 5). Consequently, the FDW contains 1,4-dioxane in the range of 1.0 µg/L, with a maximal concentration of 1.04 µg/L (Figures 4B, C, and 6, and Supporting Materials Table S2, S3).

50 km

Figure 5. Schematic overview of the research areas in Bavaria showing: Bavarian cities and municipalities whose FDW has been investigated; the main related MAR operating waterworks; and the associated rivers (Danube, Lech, Inn, Main, and Pegnitz).

In other regions of Germany, sources of 1,4-dioxane are more diffuse and require further research to be localized. For example, the FDW concentrations of 0.24 ± 0.01µg/L in the northern city districts of Berlin (supplied with RBF from Lake Tegel) suggest a contamination of the lake with 1,4-dioxane at a similar level as in the river Havel (0.24 µg/L). Also noteworthy is that the 1,4dioxane concentrations in the FDW of Thallwitz near Leipzig is almost three times higher (1.09 μg/L) than the concentration in the corresponding source water of the river Mulde (0.34 µg/L) taken on the same day, closely downstream of the RBF catchment area (Figure 4). The residence time of the water during the soil passage, and the resulting time lag between the infiltration of the river water into the aquifer and the recovery of raw water from the RBF wells might explain this discrepancy. Since only one grab sample was taken from the Mulde, it has to be considered that the high FDW value might have been caused by a concentration peak in the past that couldn’t be detected at the time of sampling. It cannot be ruled out that a more frequent point source is responsible for these concentration peaks. In general, River water concentrations of 1,4-dioxane cannot be expected to remain at the same level because of temporal dilution by rain events and variable amounts of 1,4-dioxane in discharges. Therefore, the 1,4-dioxane levels in the FDW obtained from RBF will also vary to a certain extent. Even if no significant retardation or remediation of 1,4-dioxane is expected during RBF, a depletion in 1,4-dioxane concentration in the FDW compared with the related source water is likely because of the diluting effect of the groundwater (Hamann et al., 2016; Schmidt et al., 2003; Ziegler, 2001). Groundwater from higher elevations away from the river will also contribute to the raw water in the RBF wells in varying proportions. These proportions will depend on the hydrogeological conditions at each site. Thus, the main factor influencing 1,4-dioxane contamination of the FDW obtained by MAR is the degree of pollution of the source water (Hamann et al., 2016; Rauch-Williams et al., 2010;

Schmidt et al., 2003; Schriks et al., 2010; Ziegler, 2001). A change in the condition of the surface waters, for example by elimination or addition of point pollution sources, will therefore have a direct impact on the 1,4-dioxane content of the associated FDW.

Proportion of blended water

In many cases, the analysed FDW samples comprise a mixture of water from different resources, such as MAR water blended with groundwater or unpolluted surface water from reservoirs. While the use of unpolluted surface waters is obviously preferable, it is not only the availability of unpolluted water reserves that affects decision-making in drinking water production. Blending other source waters with the local groundwater is a common practice to reduce the hardness of the FDW from groundwater. In some areas of the city of Würzburg, groundwater is not used for FDW because of its hardness. Thus, the FDW is only comprised of MAR water and consequently the amount of 1,4-dioxane in the FDW is relatively high (1.05 ± 0.02 µg/L) (Figure 6). In contrast, the FDW in the eastern part of the city of Würzburg is supplied by a mixture of MAR water and groundwater, and consequently the 14-dioxane concentration the FDW is much lower (0.3 ± 0.08 µg/L).

Figure 6. Würzburg FDW. The bars represent the mean concentration, and the error bars show the standard

deviation. /// indicates the mean of the concentrations of all FDW samples examined in Würzburg. The pure blue bar 

shows the concentration mean of the samples obtained exclusively from MAR from the river Lech; /// shows the

results of the samples obtained mainly from groundwater and blended with only small amounts of MAR water.

Similarly, the southern part of Frankfurt/Main is mainly supplied by FDW obtained from MAR systems resulting in 1,4-dioxane concentrations of 0.22 ± 0.04 µg/L, on average. The middle and northern part of Frankfurt/Main is supplied with FDW obtained from groundwater blended with various proportions of MAR water, which results in much lower 1,4-dioxane concentrations in the FDW (0.09 ± 0.04 µg/L). High standard deviations of the mean values are particularly evident in the FDW of cities where the FDW in different districts comprised different sources with varying amounts of polluted MAR water and unpolluted groundwater and/or reservoir water (Figure 7). Samples obtained from districts where the FDW at least partly consists of groundwater or reservoir water are generally less polluted than samples of FDW from adjacent districts where FDW is solely supplied by MAR systems. Therefore, samples containing a significant portion of MAR water are of particular interest for this study, although the levels of these fractions can vary greatly between sampling points, even within one city. For the sake of clarity, the results for these samples are summarized below, and are roughly divided into three groups according to the fraction of MAR water in the FDW (Figure 7).

Figure 7. Average concentrations of 1,4-dioxane in samples from districts where the presence of MAR

water in the FDW could be determined. Results of the districts that use exclusively or predominantly use groundwater for FDW supply are not shown. Bars indicate the standard deviation. The assumed proportions of MAR water in the analyzed FDW samples :

≥90%;  ≥50%;  uncertain.

When no exact

data were available, the proportion of MAR water in the FDW was deduced from the measured 1,4dioxane concentration. The colored backgrounds refer to the corresponding regions: FRANCONIA,  LEIPZIG,  MAINZ& RHINE-HESSE,  RUHR AREA,  RHINE-MAIN AREA.

 BERLIN,



In the FDW samples from Berlin, Leipzig, Dresden, Duisburg, Düsseldorf, and Mainz and the surrounding areas, the proportion of MAR water was above 50%, while the proportion of water of other origin, such as groundwater or reservoir water, ranged from 10% to 50%. For FDW samples from Würzburg, Frankfurt/Main, and Wiesbaden the high inner-city variation of the measured 1,4-dioxane concentrations was in accordance with the variance of MAR proportions in different districts. In the case of some Bavarian sites, the exact proportion of MAR in the FDW could not be determined. However, the concentrations of 1,4-dioxane in FDW from the cities of Nürnberg (max value 1.45 µg/L), Fürth (1.40 µg/L), and Erlangen (max value 2.05 µg/L), indicate that the proportion of MAR water in the FDW for certain districts of these cities is quite high.

Site-specific MAR method

Further factors that influence the levels of 1,4-dioxane in the FDW is the method of MAR (RBF vs. AGR; and the detailed treatment process), and the specific properties of individual river sections (e.g. bank/soil texture, 1,4-dioxane concentration). This is demonstrated by the case of the FDW samples supplied from MAR applied in the area of the river Rhine (Figure 4A). There was a significant difference (α = 0.05; paired t-test) in the contents of 1,4-dioxane in the FDW in Darmstadt (Ø 0.04 ± 0.01 µg/L) on the one hand and Wiesbaden (Ø 0.17 ± 0.07µg/L) and Mainz&-Surroundings (Ø 0.50 ± 0.14 µg/L) on the other hand. Even though the source water in all these cases originates from the same section of the river Rhine (Figure 4A). In Mainz and the surrounding areas, RBF from the river Rhine is used for FDW supply, while in the Hessian Ried, the application of AGR with Rhine water is the method of choice for the supply of FDW to Darmstadt and Wiesbaden. In case of the RBF implemented in Mainz and its surroundings, the sediments used for filtration consist of sand and gravel. The average sub-

surface residence time of the water during RBF through sand and gravel is 35 days (WVR, 2017). The collected water, which consists of 80%–90% of bank filtrate, is discharged to waterworks and then further processed (WVR, 2017). The FDW of Darmstadt, however, is supplied by AGR via extraction plants in the Hessian Ried. The raw water obtained there is extracted from groundwater after groundwater augmentation with pre-treated Rhine water. The pre-treatment includes, inter alia, two ozonation steps (WHR, 2017). Ozonation has previously been shown to not readily oxidize and remove 1,4-dioxane from water, but ozonation may, however, contribute to its more effective removal from Rhine water in the present case (Adams et al., 1994; Barndõk et al., 2014; Kwon et al., 2012; Lekkerkerker et al., 2009; Zenker et al., 2003). However, the key factor is probably the fundamental difference in the MAR methods used. FDW produced by AGR will generally contain a higher, but not precisely determinable, proportion of groundwater when compared with FDW obtained by RBF (up to 90% river water) (Figure 1). In the case of FDW from Darmstadt in particular, water is extracted from deep wells (40–100 m in depth) before it is further processed. The distance covered by the river water during AGR into the underground is therefore quite long, and this should have an overall positive influence on the attenuation of 1,4-dioxane by dilution. In addition, it is plausible that the proportion of natural groundwater in the extracted AGR water will ultimately be quite high. Consequently, the actual proportion of water from AGR in the groundwater of the Hessian Ried will be low, and therefore the concentrations of 1,4-dioxane in the extracted groundwater for the supply of FDW to Darmstadt will also be low. The data from Essen and Volkach also reveal the effect of the MAR method used. The FDW has the same level of 1,4-dioxane (0.11 µg/L) at both locations, whereas samples from the associated rivers show different concentrations of 1,4-dioxane (Ruhr (Essen): 0.28 µg/L; Main (Volkach):

0.11 µg/L). The FDW in each case consists of more than 90% of MAR water, although the FDW in Essen was obtained by AGR and in Volkach by RBF. In the case of the cities of Düsseldorf (Ø= 0.53 µg/L) and Duisburg (Ø= 0.18 µg/L), the abundance of 1,4-dioxane in FDW is attributable to the application of RBF in close proximity to the river Rhine. However, the standard deviation of the data indicates high fluctuations in the values for each city (Figure 4A). These fluctuations can be explained by the variation in 1,4dioxane contents in the Rhine within different catchment areas, and by different proportions of blended RBF. In Düsseldorf, contamination of groundwater with 1,4-dioxane was detected at several points. A resulting influence on the raw water in the waterworks cannot be completely ruled out. In Cologne, a total of two FDW samples were taken: one on the left bank (0.28 µg/L) and one on the right bank of the Rhine (below LOD). Only the FDW from the left bank of the Rhine was obtained from MAR water (RBF-Rhine), whereas the FDW at the right bank site consists mostly of groundwater recharged from the Bergisch Land. The FDW from the left bank originates from the Rhine and is subject to MAR. Only the concentration of the left bank sample is displayed in Figure 4A. Temporally and spatially fluctuating contamination peaks in the source waters, differing MAR methods, differences in the raw water composition, and differences in the further treatment by the individual waterworks are all critical in determining the different 1,4-dioxane concentrations (indicated by error bars) in the FDW from common source water (Figure 4C). The tested FDW, however, which were obtained by MAR, or directly, from unpolluted source waters, such as the reservoir waters of Lake Constance or Haltern/Hullen, were not contaminated with 1,4-dioxane (Figure 4C).

4 CONCLUSIONS

1,4-dioxane is a widespread FDW pollutant in those regions of Germany where MAR is used to supply FDW. This applies whether MAR uses only one source, or blending with water from other sources, such as groundwater or reservoir water. Moreover, almost every river which is exploited for FDW supply by MAR that was examined in this study, has 1,4-dioxane contamination at variable amounts at levels above the LOQ of 0.034 µg/L. Conversely, in the investigated water reservoirs and lakes from which FDW was obtained, partly directly and partly after the application of MAR, concentrations of 1,4-dioxane were below the LOQ. Consequently, in FDW obtained from these resources no 1,4-dioxane could be detected. The data generated by this study show that the 1,4-dioxane levels in the source water, the MAR methods used, as well as the proportion of blended unpolluted water into the FDW, are main parameters that determine the extent of the contamination of FDW. Based on the data, we can assume that (1) both types of MAR (RBF and AGR) are ineffective in the complete elimination of 1,4-dioxane from surface water, and (2) further treatment by waterworks does not eliminate 1,4-dioxane from the FDW. Still, the findings suggest that, with regard to the application of MAR methods, AGR is preferable to RBF. In particular, interim treatment steps are feasible or have already been implemented, and, moreover, the proportion of groundwater in the raw water obtained by AGR is usually higher than in the case of RBF. The results obtained in the present study demonstrate the close linkage between the pollution of German surface waters with 1,4-dioxane and (in the case of MAR) the resulting FDW contamination and indicate that further investigations to identify the dominant sources of 1,4-dioxane in German river systems are urgently required. Appropriate monitoring for 1,4-dioxane will help to identify and eliminate point sources of pollution by adapting the current waste water treatment methods, for example by implementation of AOPs. An

adapted blending process (using unpolluted waters) in the respective waterworks would further protect FDW quality.

ACKNOWLEDGMENTS We would like to thank all those who supported us with the sampling. Special thanks go to F. Bachmeier for her valuable support in working with Matlab. We thank P. Seward, PhD, from Edanz Group for editing a draft of this manuscript.

References Abe, A., 1999. Distribution of 1,4-dioxane in relation to possible sources in the water environment. Sci. Total Environ. 227, 41–47. https://doi.org/10.1016/S0048-9697(99)00003-0 Achten, C., Kolb, A., Püttmann, W., 2002. Occurrence of Methyl tert -Butyl Ether (MTBE) in Riverbank Filtered Water and Drinking Water Produced by Riverbank Filtration. 2. Environ. Sci. Technol. 36, 3662–3670. https://doi.org/10.1021/es011493q Adams, C.D., Scanlan, P.A., Secrist, N.D., 1994. Oxidation and Biodegradability Enhancement of 1,4Dioxane Using Hydrogen Peroxide and Ozone. Environ. Sci. Technol. 28, 1812–1818. https://doi.org/10.1021/es00060a010 Adamson, D.T., Piña, E.A., Cartwright, A.E., Rauch, S.R., Hunter Anderson, R., Mohr, T., Connor, J.A., 2017. 1,4-Dioxane drinking water occurrence data from the third unregulated contaminant monitoring rule. Sci. Total Environ. 596–597, 236–245. https://doi.org/10.1016/j.scitotenv.2017.04.085 Anderson, R.H., Anderson, J.K., Bower, P.A., 2012. Co-occurrence of 1,4-dioxane with trichloroethylene in chlorinated solvent groundwater plumes at US Air Force installations: Fact or fiction. Integr. Environ. Assess. Manag. 8, 731–737. https://doi.org/10.1002/ieam.1306 Antoniou, C.V., Koukouraki, E.E., Diamadopoulos, E., 2006. Determination of chlorinated volatile organic compounds in water and municipal wastewater using headspace–solid phase microextraction–gas chromatography. J. Chromatogr. A 1132, 310–314. https://doi.org/10.1016/j.chroma.2006.08.082 Barndõk, H., Cortijo, L., Hermosilla, D., Negro, C., Blanco, Á., 2014. Removal of 1,4-dioxane from industrial wastewaters: Routes of decomposition under different operational conditions to determine the ozone oxidation capacity. J. Hazard. Mater. 280, 340–347. https://doi.org/10.1016/j.jhazmat.2014.07.077 Barndõk, H., Hermosilla, D., Negro, C., Blanco, Á., 2018. Comparison and Predesign Cost Assessment of Different Advanced Oxidation Processes for the Treatment of 1,4-Dioxane-Containing Wastewater from the Chemical Industry. ACS Sustain. Chem. Eng. 6, 5888–5894. https://doi.org/10.1021/acssuschemeng.7b04234 BMU, 2008. Grundwasser in Deutschland, Umweltpolitik. Carrera, G., Vegué, L., Boleda, M.R., Ventura, F., 2017. Simultaneous determination of the potential carcinogen 1,4-dioxane and malodorous alkyl-1,3-dioxanes and alkyl-1,3-dioxolanes in environmental waters by solid-phase extraction and gas chromatography tandem mass spectrometry. J. Chromatogr. A 1487, 1–13. https://doi.org/10.1016/j.chroma.2017.01.015 Carrera, G., Vegué, L., Ventura, F., Hernández-Valencia, A., Devesa, R., Boleda, M.R., 2019. Dioxanes and dioxolanes in source waters: Occurrence, odor thresholds and behavior through upgraded conventional and advanced processes in a drinking water treatment plant. Water Res. 156, 404– 413. https://doi.org/10.1016/j.watres.2019.03.026 DiGuiseppi, W., Walecka-Hutchison, C., Hatton, J., 2016. 1,4-Dioxane Treatment Technologies: 1,4Dioxane Treatment Technologies. Remediat. J. 27, 71–92. https://doi.org/10.1002/rem.21498 EC (Ed.), 2002. 1,4-dioxane risk assessment ; final report. European Chemicals Agency, Helsinki. Farré, M. la, Pérez, S., Kantiani, L., Barceló, D., 2008. Fate and toxicity of emerging pollutants, their metabolites and transformation products in the aquatic environment. Adv. MS Anal. Metab. Degrad. Prod. - II 27, 991–1007. https://doi.org/10.1016/j.trac.2008.09.010 Glassmeyer, S.T., Furlong, E.T., Kolpin, D.W., Cahill, J.D., Zaugg, S.D., Werner, S.L., Meyer, M.T., Kryak, D.D., 2005. Transport of Chemical and Microbial Compounds from Known Wastewater Discharges: Potential for Use as Indicators of Human Fecal Contamination. Environ. Sci. Technol. 39, 5157–5169. https://doi.org/10.1021/es048120k Hamann, E., Stuyfzand, P.J., Greskowiak, J., Timmer, H., Massmann, G., 2016a. The fate of organic micropollutants during long-term/long-distance river bank filtration. Sci. Total Environ. 545, 629– 640. https://doi.org/10.1016/j.scitotenv.2015.12.057

Hamann, E., Stuyfzand, P.J., Greskowiak, J., Timmer, H., Massmann, G., 2016b. The fate of organic micropollutants during long-term/long-distance river bank filtration. Sci. Total Environ. 545–546, 629–640. https://doi.org/10.1016/j.scitotenv.2015.12.057 Han, T.-H., Han, J.-S., So, M.-H., Seo, J.-W., Ahn, C.-M., Min, D.H., Yoo, Y.S., Cha, D.K., Kim, C.G., 2012. The removal of 1,4-dioxane from polyester manufacturing process wastewater using an upflow Biological Aerated Filter (UBAF) packed with tire chips. J. Environ. Sci. Health Part A 47, 117–129. https://doi.org/10.1080/10934529.2012.630291 Handl, S., Mayr, E., Perfler, R., 2017. Brunnenmonitoring–Beurteilung von Einflussfaktoren auf die Leistungsfähigkeit bei Uferfiltratbrunnen. Österr. Wasser- Abfallwirtsch. 69, 240–246. Hiscock, K.M., Grischek, T., 2002. Attenuation of groundwater pollution by bank filtration. J. Hydrol. 266, 139–144. https://doi.org/10.1016/S0022-1694(02)00158-0 Isaka, K., Udagawa, M., Kimura, Y., Sei, K., Ike, M., 2016a. Biological wastewater treatment of 1,4dioxane using polyethylene glycol gel carriers entrapping Afipia sp. D1. J. Biosci. Bioeng. 121, 203–208. https://doi.org/10.1016/j.jbiosc.2015.06.006 Isaka, K., Udagawa, M., Sei, K., Ike, M., 2016b. Pilot test of biological removal of 1,4-dioxane from a chemical factory wastewater by gel carrier entrapping Afipia sp. strain D1. J. Hazard. Mater. 304, 251–258. https://doi.org/10.1016/j.jhazmat.2015.10.066 Jasmann, J.R., Borch, T., Sale, T.C., Blotevogel, J., 2016. Advanced Electrochemical Oxidation of 1,4Dioxane via Dark Catalysis by Novel Titanium Dioxide (TiO 2 ) Pellets. Environ. Sci. Technol. 50, 8817–8826. https://doi.org/10.1021/acs.est.6b02183 Jekel, M., Ruhl, A.S., Meinel, F., Zietzschmann, F., Lima, S.P., Baur, N., Wenzel, M., Gnirß, R., Sperlich, A., Dünnbier, U., Böckelmann, U., Hummelt, D., van Baar, P., Wode, F., Petersohn, D., Grummt, T., Eckhardt, A., Schulz, W., Heermann, A., Reemtsma, T., Seiwert, B., Schlittenbauer, L., Lesjean, B., Miehe, U., Remy, C., Stapf, M., Mutz, D., 2013. Anthropogenic organic micropollutants and pathogens in the urban water cycle: assessment, barriers and risk communication (ASKURIS). Environ. Sci. Eur. 25, 20. Kambhu, A., Gren, M., Tang, W., Comfort, S., Harris, C.E., 2017. Remediating 1,4-dioxane-contaminated water with slow-release persulfate and zerovalent iron. Chemosphere 175, 170–177. https://doi.org/10.1016/j.chemosphere.2017.02.044 Karges, U., Becker, J., Püttmann, W., 2018. 1,4-Dioxane pollution at contaminated groundwater sites in western Germany and its distribution within a TCE plume. Sci. Total Environ. 619–620, 712–720. https://doi.org/10.1016/j.scitotenv.2017.11.043 Krause, S., 2014. Tagungsband zum 6. Seminar Wasserversorgung: Politik, Wirtschaftlichkeit, Anlagentechnik$d, Mitteilungen / Universität der Bundeswehr, München, Institut für Wasserwesen. Shaker, Aachen. Kwon, S.C., Kim, J.Y., Yoon, S.M., Bae, W., Kang, K.S., Rhee, Y.W., 2012. Treatment characteristic of 1,4-dioxane by ozone-based advanced oxidation processes. J. Ind. Eng. Chem. 18, 1951–1955. https://doi.org/10.1016/j.jiec.2012.05.010 LDEW, 2013. Unser Trinkwasser in Hessen. Mainz. Lekkerkerker, K., Scheideler, J., Maeng, S.K., Ried, A., Verberk, J.Q.J.C., Knol, A.H., Amy, G., van Dijk, J.C., 2009. Advanced oxidation and artificial recharge: a synergistic hybrid system for removal of organic micropollutants. Water Sci. Technol. Water Supply 9, 643. https://doi.org/10.2166/ws.2009.696 LfU Bayern, 2019. 1,4-Dioxan in Bayern. LfU Bayern, 2016. PM 44/16: Environmental monitoring is being expanded - new methods are applied by the LfU (Umweltmonitoring wird ausgeweitet - LfU setzt neue Messmethodik ein). Massmann, G., Pekdeger, A., Heberer, T., Grützmacher, G., Dünnbier, U., Knappe, A., Meyer, H., Mechlinski, A., 2007. Trinkwassergewinnung in urbanen Räumen – Erkenntnisse zur Uferfiltration in Berlin. Grundwasser 12, 232–245. https://doi.org/10.1007/s00767-007-0036-7 McElroy, A.C., Hyman, M.R., Knappe, D.R.U., 2019. 1,4-Dioxane in drinking water: emerging for 40 years and still unregulated. Curr. Opin. Environ. Sci. Health 7, 117–125. https://doi.org/10.1016/j.coesh.2019.01.003

Mohr, T., Stickney, J., Diguiseppi, B., 2010. Environmental investigation and remediation: 1,4-dioxane and other solvent stabilizers. CRC Press, Boca Raton [FL]. Ortega, J.A.T., 2012. Sulfonation/sulfation processing technology for anionic surfactant manufacture, in: Advances in Chemical Engineering. InTech. Polomčić, D., Hajdin, B., Stevanović, Z., Bajić, D., Hajdin, K., 2013. Groundwater management by riverbank filtration and an infiltration channel: the case of Obrenovac, Serbia. Hydrogeol. J. 21, 1519–1530. https://doi.org/10.1007/s10040-013-1025-9 Rauch-Williams, T., Hoppe-Jones, C., Drewes, J.E., 2010. The role of organic matter in the removal of emerging trace organic chemicals during managed aquifer recharge. Water Res. 44, 449–460. https://doi.org/10.1016/j.watres.2009.08.027 Sauvé, S., Desrosiers, M., 2014. A review of what is an emerging contaminant. Chem. Cent. J. 8, 1–15. Schielein, P., Schellmann, G., 2016. Erläuterungen zur quartärgeologischen Karte 1:25.000 des Lech- und Donautals auf Blatt 7231 Genderkingen – Kartierungsergebnisse aus den Jahren 2008 und 2009. Bamberger Geographische Schriften. Schilling, C., Arznet, K., 2018. Anlage 3 Bericht der Sachverständigen Arbeitsgruppe „Bewirtschaftung und Schutz der Gewässer“.pdf. Bundeskanzleramt Österreich. Schmidt, C.K., Lange, F.T., Brauch, H.-J., Kühn, W., 2003. Experiences with riverbank filtration and infiltration in Germany. DVGW-Water Technology Center (TZW), Karlsruhe, Germany, 17. Schriks, M., Heringa, M.B., van der Kooi, M.M.E., de Voogt, P., van Wezel, A.P., 2010. Toxicological relevance of emerging contaminants for drinking water quality. Water Res. 44, 461–476. https://doi.org/10.1016/j.watres.2009.08.023 Schulze, D., 1997. Wassermengenwirtschaft im Ballungsraum Berlin. Arch Für Nat-Lands 35, 289–300. Stepien, D.K., Diehl, P., Helm, J., Thoms, A., Püttmann, W., 2014. Fate of 1,4-dioxane in the aquatic environment: From sewage to drinking water. Water Res. 48, 406–419. https://doi.org/10.1016/j.watres.2013.09.057 Stepien, D.K., Regnery, J., Merz, C., Püttmann, W., 2013. Behavior of organophosphates and hydrophilic ethers during bank filtration and their potential application as organic tracers. A field study from the Oderbruch, Germany. Sci. Total Environ. 458–460, 150–159. https://doi.org/10.1016/j.scitotenv.2013.04.020 Storck, F.R., Schmidt, C.K., Lange, F.T., Henson, J.W., Hahn, K., 2012. Factors controlling micropollutant removal during riverbank filtration. J. - Am. Water Works Assoc. 104, E643– E652. https://doi.org/10.5942/jawwa.2012.104.0147 Stuyfzand, P.J., 2016. History of managed aquifer recharge in The Netherlands. Electron. Suppl. Mater.Hydrogeol. J. Sixty Years Glob. Prog. Manag. Aquifer Recharge 36. Suh, J.H., Mohseni, M., 2004. A study on the relationship between biodegradability enhancement and oxidation of 1,4-dioxane using ozone and hydrogen peroxide. Water Res. 38, 2596–2604. https://doi.org/10.1016/j.watres.2004.03.002 Suthersan, S., Quinnan, J., Horst, J., Ross, I., Kalve, E., Bell, C., Pancras, T., 2016. Making strides in the management of “emerging contaminants.” Groundw. Monit. Remediat. 36, 15–25. U. S. EPA, 2016. 1,4-Dioxane (CASRN 123-91-1) - Report on 370 carcinogens - 14th Edition. UBA, 2016. Rund um das Trinkwasser | Umweltbundesamt. 4. Auflage. UBA, Umwelt Bundesamt, 2010. Rund um das Trinkwasser. Dessau-Roßlau. Umweltamt Düsseldorf, 2017. 19_Broschuere_Grundwassersanierung_bf.pdf [WWW Document]. URL https://www.duesseldorf.de/fileadmin/Amt19/umweltamt/publikationen/19_Broschuere_Grundwa ssersanierung_bf.pdf (accessed 1.30.17). UNEP, 2006. Handbook for the Montreal protocol on substances that deplete the ozone layer. UNEP/Earthprint. U.S. EPA, 2008. EPA Report 600-R-08-101. METHOD 522 DETERMINATION OF 1, 4-DIOXANE IN DRINKING WATER BY SOLID PHASE EXTRACTION (SPE) AND GAS CHROMATOGRAPHY/MASS SPECTROMETRY (GC/MS) WITH SELECTED ION MONITORING (SIM). US EPA (US Environmental Protection Agency).

US EPA, O., 2018. 2018 Drinking Water Standards and Advisory Tables [WWW Document]. US EPA. URL https://www.epa.gov/dwstandardsregulations/2018-drinking-water-standards-and-advisorytables (accessed 4.10.19). US-EPA, 2013. Technical Fact Sheet – 1,4-Dioxane | Cleanups at Federal Facilities | US EPA [WWW Document]. URL https://www.epa.gov/fedfac/technical-fact-sheet-14-dioxane (accessed 6.8.16). US-EPA IRIS, 2010. TOXICOLOGICAL REVIEW OF 1,4-Dioxane.pdf In Support of Summary Information on the Integrated Risk Information System (IRIS) [WWW Document]. URL http://cfpub.epa.gov/ncea/iris/iris_documents/documents/toxreviews/0326tr.pdf (accessed 6.8.16). VEBW, Hövelborn, J., Fischer, D., 2013. Wasser für Bayern. München. Vescovi, T., Coleman, H.M., Amal, R., 2010. The effect of pH on UV-based advanced oxidation technologies – 1,4-Dioxane degradation. J. Hazard. Mater. 182, 75–79. https://doi.org/10.1016/j.jhazmat.2010.06.001 Wakayama, H., 2003. Revision of Drinking Water Standards in Japan [WWW Document]. URL http://www.nilim.go.jp/lab/bcg/siryou/tnn/tnn0264pdf/ks0264011.pdf (accessed 11.2.17). Weigelhofer, G., Reckendorfer, W., Funk, A., Hein, T., 2013. Auenrevitalisierung–Potenzial und Grenzen am Beispiel der Lobau, Nationalpark Donau-Auen. Österr. Wasser- Abfallwirtsch. 65, 400–407. WHO (Ed.), 2011. Guidelines for drinking-water quality, 4th ed. ed. World Health Organization, Geneva. WHR, 2018. Das Hessische Ried zwischen Vernässung und Trockenheit: eine komplexe wasserwirtschaftliche Problematik. WHR, 2017. wasserverband Hessisches Ried [WWW Document]. URL http://www.wasserverbandhessisches-ried.de/html/wasseraufbereitung.html (accessed 7.18.17). Woodard, S., Mohr, T., Nickelsen, M.G., 2014. Synthetic Media: A Promising New Treatment Technology for 1,4-Dioxane: A Promising New Treatment Technology for 1,4-Dioxane. Remediat. J. 24, 27–40. https://doi.org/10.1002/rem.21402 WVR, 2017. Uferfiltratwerk Bodenheim - Wasserversorgung Rheinhessen Pfalz [WWW Document]. URL https://www.wvr.de/wasserinfo/wasserwerke/70-uferfiltratwerk-bodenheim (accessed 7.17.17). Zenker, M.J., Borden, R.C., Barlaz, M.A., 2003. Occurrence and Treatment of 1,4-Dioxane in Aqueous Environments. Environ. Eng. Sci. 20, 423–432. https://doi.org/10.1089/109287503768335913 Zhang, S., Gedalanga, P.B., Mahendra, S., 2017. Advances in bioremediation of 1,4-dioxane-contaminated waters. J. Environ. Manage. 204, 765–774. https://doi.org/10.1016/j.jenvman.2017.05.033 Ziegler, D.H., 2001. Untersuchungen zur nachhaltigen Wirkung der Uferfiltration im Wasserkreislauf Berlins. Technische Universität Berlin. https://doi.org/10.14279/depositonce-419

Final drinking water, obtained by MAR, was tested for 1,4-dioxane throughout Germany. Managed Aquifer Recharge Systems (MAR) are decisive for 1,4-dioxane in German FDW. 98 out of 125 drinking water samples contained 1,4-dioxane of 0.034 to 2.05 μg/L. Source water pollution, blending and exact method of MAR are key factors identified. Bank filtration exhibits to be less favorable than artificial groundwater recharge.

Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: