Rare earth elements as emerging contaminants in the Rhine River, Germany and its tributaries

by

Serkan Kulaksız

A thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy in Geochemistry

Approved, Thesis Committee ______Prof. Dr. Michael Bau, Chair Jacobs University Bremen ______Prof. Dr. Andrea Koschinsky Jacobs University Bremen ______Dr. Dieter Garbe-Schönberg Universität Kiel

Date of Defense: June 7th, 2012 ______School of Engineering and Science

TABLE OF CONTENTS CHAPTER I – INTRODUCTION 1

1. Outline 1 2. Research Goals 4 3. Geochemistry of the Rare Earth Elements 6 3.1 Controls on Rare Earth Elements in River Waters 6 3.2 Rare Earth Elements in Estuaries and Seawater 8 3.3 Anthropogenic Gadolinium 9 3.3.1 Controls on Anthropogenic Gadolinium 10 4. Demand for Rare Earth Elements 12 5 Rare Earth Element Toxicity 16 6. Study Area 17 7. References 19 Acknowledgements 28

CHAPTER II – SAMPLING AND METHODS 31

1. Sample Preparation 31 1.1 Pre‐concentration 32 2. Methods 34

2.1 HCO3 titration 34 2.2 Ion Chromatography 34 2.3 Inductively Coupled Plasma – Optical Emission Spectrometer 35 2.4 Inductively Coupled Plasma – Mass Spectrometer 35 2.4.1 Method reliability 36 3. References 41

CHAPTER III – RARE EARTH ELEMENTS IN THE RHINE RIVER, GERMANY: FIRST CASE OF ANTHROPOGENIC LANTHANUM AS A DISSOLVED MICROCONTAMINANT IN THE HYDROSPHERE 43

Abstract 44 1. Introduction 44 2. Sampling sites and Methods 46 2.1 Samples 46 2.2 Methods 46 2.3 Quantification of REE anomalies 47 3. Results and Discussion 48 4. Conclusions 49 Acknowledgements 49 5. References 49 CHAPTER IV – NATURAL AND ANTHROPOGENIC RARE EARTH ELEMENT DISTRIBUTION OF THE RHINE RIVER AND ITS TRIBUTARIES 51

1. Introduction 53 2. Sampling and Methods 58 2.1 Sample Processing and Analysis 58 2.2 Quantification of anomalies / Extrapolation of missing data 59 2.3 Study Area 60 2.3.1 Rhine River Sections 61 2.3.2 Lake Constance 62 3. Results 63 3.1 Major Ions 64 3.1.1 Rhine River 64 3.1.2 The effect of the input of effluent at Rhine‐km 447.3 75 3.1.3 Tributaries 77 3.1.3.1 Aare 80 3.1.3.2 Main 80 3.1.3.3 Mosel 80 3.1.3.4 Neckar 81 3.1.3.5 Lippe 81 3.1.3.6 Wupper 82 3.2 Rare Earth Elements 82 3.2.1 Rare Earth Elements in the Rhine River 82 3.2.2 Rare Earth Elements in the Tributaries 86 3.2.2.1 Aare 87 3.2.2.2 Main 88 3.2.2.3 Mosel 89 3.2.2.4 Neckar 90 3.2.2.5 Lippe 90 3.2.2.6 Wupper 91 4. Discussion 93 4.1 Natural Rare Earth Elements 93 4.2 Ce Anomaly 94 4.3 Anthropogenic Gd 98 4.3.1 Variation of anthropogenic Gd with weekday 101 4.4 Anthropogenic La 105 4.5 Anthropogenic vs. Natural Rare Earth Elements 107 4.5.1 Anthropogenic La vs. Gd 109 5. Conclusions and Outlook 117 5.1 Rare Earth Elements as Emerging Contaminants 117 5.2 Suggestions and Outlook 118 Acknowledgements 120 6. References 121 CHAPTER V – ANTHROPOGENIC DISSOLVED AND COLLOID/NANOPARTICLE‐BOUND SAMARIUM, LANTHANUM AND GADOLINIUM IN THE RHINE RIVER AND THE IMPENDING DESTRUCTION OF THE NATURAL RARE EARTH ELEMENT DISTRIBUTION IN RIVERS 129

Abstract 130 1. Introduction 130 2. Methods 131 2.1. Sampling and analysis 131 2.2 Quantification of REE anomalies 133 3. Results 134 3.1 Dissolved REE 134 3.2. The truly dissolved and nanoparticulate REE pools 134 4. Discussion 135 4.1 Samarium as another emerging REE microcontaminant 135 4.2 Truly dissolved and nanoparticulate anthropogenic REE 135 4.3 Quantifying anthropogenic REE transport via the Rhine River 135 4.4 Environmental impact of anthropogenic REE 136 5. Conclusions 136 Acknowledgements 136 References 136

CHAPTER VI ‐ ANTHROPOGENIC GADOLINIUM AS A MICROCONTAMINANT IN TAP WATER USED AS DRINKING WATER IN URBAN AREAS AND MEGACITIES 139

Abstract 140 1. Introduction 140 2. Materials and Methods 142 2.1 Sampling 142 2.2 Quantification of anomalies 142 3. Results and Discussion 144 3.1 Distribution of uranium, barium, rubidium and strontium 144 3.2 Rare earth elements 144 3.2 Toxicity of Gd3+ and Gd‐CA 146 4. Conclusions 147 Acknowledgements 147 5. References 147

CHAPTER VII – CONCLUSIONS 151

1. Documenting and quantifying anthropogenic Rare Earth Elements 151 1.1 Anthropogenic Gd 151 1.2 Anthropogenic La 151 1.3 Anthropogenic Sm 152 2. Controls on Rare Earth Elements in the Rhine River Catchment 152 2.1 Natural Rare Earth Elements 152 2.2 Anthropogenic Rare Earth Elements 153 3. The Role of Ultrafiltration 153 4. Rare Earth Elements as Emerging Contaminants in Tap Water 154 5. Closing Remarks 156 APPENDIX 1 159 Shale normalized REE patterns of the Rhine River

APPENDIX 2 167 Shale normalized REE patterns of the tributaries of the Rhine River

APPENDIX 3 174 Data for the Rhine River in May 2008

APPENDIX 4 179 Data for the Rhine River in May 2009

APPENDIX 5 184 Data for the Rhine River in October 2009

APPENDIX 6 189 Data for the tributaries of the Rhine River in May 2008

APPENDIX 7 194 Data for the tributaries of the Rhine River in May 2009

APPENDIX 8 199 Data for the tributaries of the Rhine River in October 2009

APPENDIX 9 204 Data for the Wiembach Creek and Lake Constance

APPENDIX 10 206 General information and references for the tributaries of the Rhine River

APPENDIX 11 208 Calculation of Rare Earth Element Anomalies

Andreas Gursky’s $4.3m photo (Rhein II, 1999) of the Rhine River1:

„The Rhine is the river about which all the world speaks but no one studies, which all the world visits but no one knows, which one sees as it passes but forgets as it flows, which everyone skims but no one plumbs. Still, its ruins lift the imagination and its destiny preoccupies serious minds; and below the surface of its current, this admirable river reveals to the poet and statesmen alike the past and future of Europe.“

Victor Hugo, 1845

1 Fair use rationale: The image is a low resolution copy of the original work, and of such low quality that it will not affect potential sales of the art work.

CHAPTER I – INTRODUCTION

1. OUTLINE

This dissertation is divided into 7 chapters, with each one focusing on a specific aspect of the general research questions. The first chapter states the objectives of this study and provides an introduction into rare earth elements (REE) in aqueous media. General information on the Rhine River and its tributaries is provided, as well as a map showing sampling locations.

Analytical methods are presented in the short Chapter II, including sample preparation, workflows and assessment of the analytical quality of the dataset. Measured elements include

‐ ‐ ‐ 2‐ + + 2+ 2+ ‐ major anions (Cl , Br , NO3 and SO4 ), major cations (Na , K , Mg and Ca ), Si, HCO3 , trace elements such as Ba, Rb, Sr, U, Y and REE (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu).

Chapter III presents one of the key findings in this study: the first report of anthropogenic La anomalies in rivers worldwide is reported. The source of the La contamination is presented, and possible implications for the environment and natural REE studies are discussed. The La contamination is traced to a point source at Rhine‐km 447.3: effluent discharge from a factory producing fluid catalytic cracking (FCC) catalysts. In the vicinity of the effluent pipe, La concentrations of up to 49 mg/kg are measured, well above ecotoxicological levels. This chapter was published in Vol. 37 of the journal Environment International in 2011 with the title “Rare earth elements in the Rhine River, Germany: First case of anthropogenic lanthanum as a dissolved microcontaminant in the hydrosphere” and can be found online at http://dx.doi.org/10.1016/j.envint.2011.02.018.

Chapter IV presents largely unpublished data and represents the first extensive REE dataset on the Rhine River and its tributaries. Sampling was carried out in May 2008, May 2009 and October 2009 covering the Rhine River from the Swiss – German border to the German – Dutch border. Natural REE and anthropogenic REE are discussed separately, with emphasis on anthropogenic REE. Anthropogenic Gd increases steadily going downstream due to its diffuse source nature, while anthropogenic La emerges at approximately Rhine‐km 450 due to its point‐source nature

1 and is then diluted further downstream. The effect of seasonal discharge is discussed with regards to natural and anthropogenic REE. Several tributaries in the Ruhr‐area (a conglomerate megacity in the Middle Rhine) show exceptionally high Gd anomalies due to the increased amount of anthropogenic Gd input and the small discharge of the tributaries. A tendency for reduced Gd anomalies on the first weekdays can also be observed. Based on discharge and concentration data, estimates of yearly export of anthropogenic Gd and La to the North Sea via the Rhine River are given (Gd: 53 – 215 kg/year; La: 862 – 1090 kg/year). This chapter is titled “Natural and Anthropogenic Rare Earth Element Distribution of the Rhine River and its Tributaries” and is in preparation for submission.

Chapter V reports on the first occurrence of anthropogenic Sm anomalies in rivers worldwide. The Rhine River has Sm concentrations up to 7.7 times higher than background concentrations. The Sm contamination in the Rhine River has emerged between October 2010 and May 2011 and its source is the effluent pipe of the same facility that produces FCC catalysts. Ultrafiltration data that shows the truly dissolved nature of anthropogenic Gd is also provided. In contrast, anthropogenic La and anthropogenic Sm are associated with the nanoparticulate REE pool as well as the truly dissolved REE pool. Based on data presented in this chapter, the Rhine River alone carries between 330 – 5700 kg of anthropogenic La, 146 – 584 kg of anthropogenic Sm, and 329 – 730 kg of anthropogenic Gd toward the North Sea each year. This chapter is titled “Anthropogenic dissolved and colloid/nanoparticle‐bound samarium, lanthanum and gadolinium in the Rhine River and the impending destruction of the natural rare earth element distribution in rivers” and has been published in Vol. 362 of the journal Earth and Planetary Science Letters in 2013 and can be found online at http://dx.doi.org/10.1016/j.epsl.2012.11.033.

Chapter VI presents dissolved REE data in tap water from the city of Berlin, where anthropogenic Gd was first reported in the small rivers receiving large amounts of wastewater treatment plant (WWTP) effluent. The groundwater management of the former west part of Berlin historically relied (and still does) on artificial and natural groundwater recharge to replenish their sources of drink water. In contrast, the former east part of Berlin did not artificially recharge groundwater. Data from each borough of the city yields a remarkable pattern, in which tap water from the eastern boroughs show no anthropogenic Gd contamination, while tap water from the western boroughs show significantly high concentrations of anthropogenic Gd. The sample from the

2 German Parliament (Reichstag) showed one of the largest Gd anomalies in the dataset. This chapter was published in Vol. 26 of the journal Applied Geochemistry in 2011 with the title “Anthropogenic gadolinium as a microcontaminant in tap water used as drinking water in urban areas and megacities” and can be found online at http://dx.doi.org/10.1016/j.apgeochem.2011.06.011.

Finally, Chapter VII summarizes the results presented here and offers a general conclusion. Open questions and ideas for future studies are discussed. Among the suggestions are long‐term monitoring of REE concentrations and distribution at a given location to better our understanding of the behavior of natural and anthropogenic REE as well as a detailed study of the Rhine‐Meuse estuary for investigating the presence and export of anthropogenic La, Sm and Gd into the North Sea. A large portion of the population living within the Rhine River catchment relies on the Rhine River or its tributaries for drinking water. A detailed study of REE in cities that depend on the Rhine River and other European cities / rivers is proposed, similar to the study of tap water in Berlin (Chapter V).

As several REE show anomalous behavior (natural and anthropogenic), Appendix XI also describes calculations of background REE concentrations and REE anomalies. A new technique for calculating REE anomalies is suggested, in order to provide an alternative to previously used but often unsuitable methods in literature.

3 2. RESEARCH GOALS

The rare earth elements (REE) are ideal as geochemical tracers of natural processes in pristine environments due to their coherent behavior as a group. However, studies of this nature are compromised since the first report of anthropogenic REE in surface waters in 1996 (Bau and Dulski 1996): Anthropogenic gadolinium (Gd) used as contrast agents for computer tomography is reported in small rivers in the Berlin area as well as the Rhine River and its tributary Wupper River. Since then, numerous rivers with anthropogenic Gd contamination have been reported worldwide.

The Rhine River also carries anthropogenic Gd anomalies, as well as a more recently reported anthropogenic lanthanum (La) anomaly, locating the source of which was one of the key aims of this study. The rapid speed with which new anthropogenic REE appear as emerging contaminants in surface water is alarming. Anthropogenic La and anthropogenic Gd dominate dissolved REE concentrations (e.g. 97%) and influence their distribution severely in the Rhine River. The presence of anthropogenic REE in surface waters not only hampers studies of natural REE as geochemical tracers, but also poses a potential ecotoxicological threat especially in the absence of detailed studies that focus on long‐term low‐dosage exposure and synergetic effects.

Currently no extensive REE dataset that includes the whole Rhine River and its tributaries is available to serve as background levels for monitoring studies. With the rate of increase in REE applications and the demand for REE worldwide, these monitoring studies and background levels will be important in determining legislature and policy making regarding legal limits in the environment. At present legal limits for REE largely do not exist. With the ongoing struggle for securing REE resources, and the concern for the future of affordable REE supplies, the environmental aspects are not the priority for industrialized countries. This is why the dataset presented here and others with similar aims are crucial for establishing base levels.

Although REE distribution in large river systems such as the Amazon River (Sholkovitz, 1993; Gaillardet et al., 2003; Gerard et al., 2003) and the Mississippi River (Shiller, 2002) have been studied extensively, no similar study has been undertaken for the Rhine River. The distribution of natural REE for the Rhine River and its tributaries are largely unknown and variation in natural

4 and anthropogenic REE between different discharge periods has not been investigated. Moreover, an effort to determine different sources and behavior of natural and anthropogenic REE on a scale this large has not been made. More specifically, the effect of discharge and other factors that influence anthropogenic REE is lacking. For Gd, the population density of the catchment area, sampling day of the week and dilution due to increased discharge and increased natural REE levels are the predicted controlling factors. For La, the important factors are expected to be: input from the point source and dilution due to runoff and tributaries converging with the Rhine River. Mass balance calculations and flux of REE via the tributaries into the Rhine River and via the Rhine River towards the North Sea were not well constrained. The goal of this thesis is to shed light on some of these issues and provide a baseline for future studies to build on.

5 3. GEOCHEMISTRY OF THE RARE EARTH ELEMENTS

In addition to the great economic interest that REE receive and their crucial importance for numerous high‐tech products and processes, REE are also widely used as valuable tools in geochemical research due to their coherent behavior in surface waters. REE are exclusively trivalent in low‐temperature systems (with the exception of cerium, Ce), and their ionic radii REE decrease systematically with increasing atomic number due to the progressive filling up of the 4f orbital (lanthanide contraction). Hence, REE behave coherently in geochemical systems and REE concentrations reveal smooth patterns when normalized to a natural reference system (chondrite or shale) and plotted on a semi‐logarithmic scale. However, small but systematic changes in REE behavior can be observed due to differences between REE stability constants for surface and solution complexes or between REE mineral/melt partition coefficients. The interplay between solution and surface complexation is an important factor on the distribution of REE. Depletion of light REE (LREE) or heavy REE (HREE) may reveal information about the sources or speciation of REE in natural systems such as adsorption onto organic substances and dissolved complex formation. A change in the electron valence of REE can lead to the fractionation of the individual element from the otherwise smooth normalized pattern. Natural Ce and Eu anomalies are observed due to the redox chemistry of these elements. Cerium(III) can be oxidatively scavenged, oxidized to Ce(IV) inorganically or via microbial activity, depleting the dissolved phase of Ce(III), resulting in negative Ce anomalies in the dissolved phase. Positive Eu anomalies may be observed in high‐temperature low‐redox environments (e.g. hydrothermally influenced waters) due to the reduction of Eu(III) to Eu(II). Other anomalies, usually only found in seawater, can be observed due to small differences in solution complexation stabilities of the individual REE, e.g. La, Gd, lutetium (Byrne and Kim, 1990; Kim et al., 1991; de Baar et al., 1991). Due to the similar ionic radii and identical surface charge of yttrium (Y) and holmium (Ho), Y has been often plotted together with REE (Bau, 1996) in REE and Y (REY) plots. Despite being so‐ called “geochemical twins”, Y can be decoupled from Ho and other neighboring REE to produce positive Y anomalies due to numerous processes in aqueous systems or source rock composition in catchment areas.

3.1 CONTROLS ON RARE EARTH ELEMENTS IN RIVER WATERS

Elderfield et al. (1990) have suggested that REE are found in three distinct pools in natural waters, a concept still widely used and relevant today. An operational filter size is defined to

6 separate the particulate pool from the so‐called dissolved pool. Typical filter sizes of 0.2 µm or 0.45 µm have been commonly used, although different sizes have also been used for different investigations. The filtrate consists of the truly dissolved pool (free ions and solution complexes) and the colloidal pool (nanoparticles). Due to the unavailability of reliable methods in the past for separation of the two pools, the differentiation between dissolved and truly dissolved has only become more relevant recently. REE in the truly dissolved pool is highly bioavailable, especially “free” REE ions, and speciation of truly dissolved REE can be thermodynamically modeled. The colloidal pool is dominated by colloids, which in turn control REE concentrations in the other pools: distribution in the dissolved pool due to their high REE content, and the truly dissolved fraction due to their high scavenging capacity. The latter point is due to the particle‐ reactivity of REE and their strong association with colloids.

The colloidal pool is not easily removed from the truly dissolved pool due to the unavailability of reliable and cost‐efficient removal methods. Methods such as ultrafiltration, ultracentrifugation or thin film techniques, are costly, time consuming, and might impose artifacts e.g. truly dissolved elements may be retained inside the filtration system (Sholkovitz, 1995; Horowitz et al., 1996; Hoffmann et al., 2000). The ultrafiltration technique most often employed, cross‐flow ultrafiltration (CFF), is operator and equipment dependent (Buesseler et al., 1996) and causes fractionation of colloidal components (Gustafsson et al., 1996). Therefore, REE data is commonly reported as the “dissolved” concentration, while in reality the sum of the truly dissolved concentration and the colloidal concentration is measured. Nevertheless, ultrafiltration and ultracentrifugation are necessary in order to gain insight into the role of colloids and the partitioning of REE between the colloidal and truly dissolved fraction. Dupré et al. (1999) have performed ultrafiltration on organic‐rich river waters with successively smaller cut off sizes, and reported progressively decreasing trace element concentrations in each filtrate. Pokrovski and Schott (2002) report two different types of colloids (organic colloids and iron oxyhydroxides) in boreal organic‐rich rivers and have shown the association of organic colloids with HREE while the larger iron oxyhydroxides colloids generally associate with REE.

Dissolved organic carbon (DOC) and pH are two of the most important factors controlling REE in river waters. Dissolved REE concentrations decrease with increasing pH, a parameter controlled mainly by organic substances present in the river. The close association of REE and DOC in rivers

7 has been extensively documented (Sholkovitz, 1978; Sholkovitz et al., 1978; Elderfield et al., 1990; Ingri, et al., 2000). Especially in carbonate rivers, the relationship of decreasing REE concentrations with increasing pH is well documented (Elderfield et al., 1990).

Shale‐normalized REE patterns in river waters (reviewed by, e.g., Elderfield et al., 1990; Sholkovitz, 1995; Gaillardet et al., 2003) show a large variation ranging from light REE (LREE) enriched to heavy REE (HREE) enriched. Changes in the relative abundance of truly dissolved and colloidal pools in different settings intertwined with the role of DOC and pH act collectively to this effect. A simplified two‐component mixing between the colloidal and truly dissolved pools (Elderfield et al., 1990) can account for the differences observed in REE distribution. Colloids carry up to > 90 % of REE in rivers with high dissolved organic carbon (DOC), and are characterized by LREE‐enriched REE patterns compared to the truly dissolved REE pool. In contrast, the truly dissolved load is HREE‐enriched and constitutes a minor fraction of the total REE. The higher the amount of colloids in a sample, the higher the REE content, and the less HREE‐enriched the shale normalized REE pattern (Elderfield et al., 1990; Sholkovitz, 1995).

3.2 RARE EARTH ELEMENTS IN ESTUARIES AND SEAWATER

The distribution of truly dissolved and colloidal REE in rivers affects REE distribution in seawater via processes in estuaries. As seawater is admixed into river water, the ionic strength of the solution changes, altering the surface charges of colloids. As a result, electrostatic repulsion between colloids decreases, leading to the coagulation and removal of colloids from solution. This process is known as “salting‐out” and is most effective in the low‐salinity region of estuaries with LREE being effected the most (Hoyle et al., 1984; Goldstein and Jacobsen, 1988; Elderfield et al., 1990; Sholkovitz., 1992; Sholkovitz, 1993; Sholkovitz, 1995; Nozaki et al., 2000; Sholkovitz and Szymczak, 2000; Lawrence and Kamber, 2006) The removal of colloids from solution is accompanied by a decrease in the associated Fe‐oxyhydroxides, DOC, and particle‐reactive elements (e.g. REE) from solution. As the colloidal pool is more LREE enriched compared to the truly dissolved fraction, this process also removes LREE preferentially, leaving the truly dissolved pool enriched in HREE in solution. At higher salinity, an increase in dissolved REE concentrations is observed possibly as a result of remineralization and release of REE from particles. The general REE pattern of seawater is a result mainly of these two processes in estuaries (Sholkovitz, 1992).

8 The general REE signature of seawater is well defined and is characterized by HREE enrichment over the LREE and a number of anomalies for Ce, La, Gd and possibly Lu. The enrichment of HREE

2‐ results from the exchange equilibrium between solution‐complexation of the REE with CO3 and surface‐complexation with OH‐ and/or organic functional groups on particle surfaces. Negative Ce anomalies are reported in all oxic marine samples due to the presence of Ce in a tetravalent state as a result of oxidation. The marine Y/Ho ratio is significantly higher than that of river waters, displayed as a positive Y anomaly with super‐chondritic Y/Ho ratios. While the theoretical background for La, Gd (De Baar et al., 1985; Bau et al., 1997; Zhang and Nozaki, 1998), Lu and Y anomalies involves subtle differences between the stabilities of REY(III) complexes, the presence of a small positive Lu anomaly is still a matter of debate (De Baar et al., 1991; Bau, 1999; Bau and Alexander, 2006).

3.3 ANTHROPOGENIC GADOLINIUM

Excluding the (small) natural (La, Gd, Lu) anomalies that may be found in seawater, and anomalies of redox sensitive REE (Ce and Eu) in river waters, there are no known natural processes or sources that could produce significant fractionation of single REE. However, Bau and Dulski (1996) reported anomalously high concentrations of Gd in the low‐discharge Havel River, downstream of Berlin, as well as the Rhine River and its tributaries Wupper and Sieg. The large anthropogenic Gd anomalies were traced to Gd‐based contrast agents (Gd‐CA) such as Gd‐ DTPA and its derivatives used in magnetic resonance imaging. As these compounds are administered intravenously, they are designed for high stability and low bioavailability and are water‐soluble. The combination of these properties increases their persistence in the aqueous environment, enabling them to pass unhindered through WWTP processes. Consequently, surface waters that receive discharge from WWTP are contaminated with anthropogenic Gd (Bau and Dulski, 1996; Möller et al., 2002; Verplanck et al., 2005; 2010; Bau et al., 2006; Kulaksız and Bau, 2007; Lawrence et al., 2009; Petelet‐Giraud et al., 2009; Rabiet et al., 2009). As a result of natural and artificial groundwater recharge, groundwater (Möller et al., 2000; Knappe et al., 2005; Rabiet et al., 2006) and tap water (Bau and Dulski, 1996; Möller et al., 2002) show anthropogenic Gd anomalies. There is a growing body of literature on anthropogenic Gd in rivers and lakes worldwide (Tricca et al., 1999; Nozaki et al., 2000; Elbaz‐Poulichet et al., 2002; Möller et al., 2002, 2003; Knappe et al., 2005; Rabiet et al., 2005, 2009; Zhu et al., 2004, 2005,

9 Verplanck et al., 2005; Bau et al., 2006; Kulaksız and Bau, 2007; Lawrence et al., 2006; Morteani et al., 2006; Petelet‐Giraud et al., 2009). Anthropogenic Gd is unaffected by the removal of colloids in estuaries, as it is not associated with colloids and stays in solution, leading to anthropogenic Gd anomalies in estuaries and coastal waters (Elbaz‐Poulichet, 2002; Zhu et al., 2004; Kulaksız and Bau, 2007; Lawrence, 2010).

3.3.1 Controls on Anthropogenic Gadolinium

Besides the list of studies documenting the occurrence of anthropogenic Gd in the previous section, several authors have investigated detailed aspects of anthropogenic Gd such as behavior and half‐life in the environment. Kümmerer and Helmers (2000) have identified hospital effluents as a source of anthropogenic Gd in the hydrosphere while Künnemeyer et al. (2009) have studied the Gd‐CA in hospital and WWTP effluents and verified the speciation of anthropogenic Gd as Gd‐CA. Verplanck et al. (2010) have characterized the behavior of anthropogenic Gd through large WWTPs in the USA and concluded no removal of anthropogenic Gd through WWTP processes. Lawrence et al. (2010) have examined the removal of Gd‐CA through so‐called “advanced treatment plants” that utilize reverse osmosis, and report up to 99% removal of Gd‐CA.

Anthropogenic Gd has been demonstrated as an ideal tracer with its long environmental half‐life and unrestricted exchange between different compartments of the hydrosphere. It is a cost‐ efficient and sensitive proxy for other emerging (micro)contaminants that are potentially more harmful in the environmental but more difficult to detect. Möller et al. (2000) have successfully applied anthropogenic Gd as a conservative tracer in surface and ground water from Berlin and Lawrence (2010) has shown the ability of anthropogenic Gd as a tracer of effluent plume dispersion in a study of the Brisbane River and Moreton Bay that receives WWTP effluent. Morteani et al. (2006) used anthropogenic Gd as a proxy for WWTP effluent in their study of anthropogenic estrogen in the hydrological basin of Prague. Möller et al. (2010a; 2010b) have shown that transmetallation of Gd‐DTPA occurs with other REE as well as Y and Cu, and estimate 70 years as the time required before equilibrium is reached. Investigating the behavior of Gd‐ DTPA in an experimental setup simulating bank filtration, Möller et al. (2011) suggested that microbial degradation of Gd‐CA is not significant.

10 Large positive anthropogenic Gd anomalies are fundamentally linked to rivers within densely populated catchments in countries with highly developed healthcare systems (Bau et al., 2006). This is precisely the situation with the Rhine River and its tributaries, especially in the Middle and Lower Rhine sections. The Rhine River receives waste water treatment plant (WWTP) effluents of 96% of the inhabitants of its catchment area (ICPR, 2005) and anthropogenic Gd anomalies in the Rhine River have already been reported by Bau and Dulski (1996) and Tricca et al. (1999).

11 4. DEMAND FOR RARE EARTH ELEMENTS

Rare earth elements are considered exotic elements. However they are not rare, both in terms of their abundance in nature and the number of applications they are used in technologies embedded in our every‐day lives. Some REE are more abundant in the earth’s crust than copper, lead, gold and platinum, and they find applications in products and processes essential for our high‐tech civilization, often with no known substitute. Some examples of applications of the REE are given below and summarized in Table 1.

Since the early 1970s, REE have been used in agriculture for increasing crop yields (Brown et al., 1990; Asher, 1991; Xiong, 1995; Wang et al., 2001). REE were shown to improve the bioavailability of Ca and Mn in soil (Chang, 1991), to stimulate the synthesis of chlorophyll (Guo, 1988), to promote seedling development (Wu et al., 1983; Chang, 1991) and root and shoot growth in crops such as wheat and corn (Wu et al., 1983; Wu et al., 1985). More interestingly, REE have been used in animal husbandry (poultry and pigs) with the result of up to 20% gain in body weight in piglets (Redling, 2006, and references therein).

Lanthanum is used in mixtures with clay for removing filterable reactive phosphorus (FRP) in lakes and rivers where the aquatic systems are degraded by toxic algal blooms. For example, Phoslock®, a commercial product (National Industrial Chemicals Notification and Assessment Scheme, 2001; Phoslock Water Solutions Limited, 2011), is obtained by mixing bentonite with La which is adsorbed onto the clay and then scavenges the FRP present in the water. The result is a lanthanum phosphate, which is insoluble and forms a stable layer of “sediment” unless physico‐ chemical conditions change.

Yttrium‐stabilized zirconia and thoria are candidates for the reduction of the long‐term radiotoxicity of nuclear waste during disposal through the fixation of long‐lived actinides as target material for transmutation and as stable materials for long‐term final disposal.

Radioactive Samarium‐153 is used in medicine to treat the severe pain associated with cancers that have spread to the bone. Quadramet®, a commercial drug, is an oncology therapeutic agent

12 Table 1 – Applications of rare earth elements. Modified after Humphries, 2011.

La Energy efficient fluorescent lamps, rechargeable batteries, studio lighting and projection, Camera and telescope lenses, doping metals for change in physical properties, thermionic electron emission sources for scanning electron microscopes, petroleum cracking catalysts, medicinal treatment for hyperphosphatemia, electron‐ dense tracer in molecular biology

Ce Polishing agent for glass, energy efficient fluorescent lamps, catalytic converters in engines, petroleum cracking catalyst, doping metals for change in physical properties

Pr Permanent magnet technology, hard disk and DVD drives, doping metals for change in physical properties, studio lighting and projector lights, magnetocaloric effect (to reach very close to absolute zero), slowing down light

Nd Permanent magnet technology, hard disk and DVD drives, regenerative breaking systems, audio equipment (microphones, speakers etc.), gain media in lasers

Pm Beta radiation source for thickness gauges, luminous paint applications, nuclear battery

Sm Masers/lasers, neutron absorber, permanent magnets, treatment for bone cancer

Eu Color cathode‐ray tubes, phosphors in CRT, LCD, PTV, energy efficient fluorescent lamps, lasers, screening for some genetic diseases, anti‐counterfeiting phosphors (Euro notes)

Gd Phosphors in TV tubes, CDs, RAM, burnable poison in nuclear submarines, emergency shutdown mechanism in nuclear reactors, metallurgy, intravenous radiocontrast agent, doped in phosphor layer of X‐ray detectors, scintillator in PET scans, neutron radiography, energy efficient fluorescent lamps, magnetic refrigeration

Tb Energy efficient fluorescent lamps, glass and phosphors in CRT, LCD, PTV, doping in solid state material

Dy Permanent magnet technology, hard disk and DVD drives, lasers, nuclear control rods, infrared radiation source, contrast agent in MRI imaging, nanomagnets

Y Energy efficient fluorescent lamps, glass and phosphors in CRT, LCD, PTV, microwave filters, acoustic energy transmitter and transducer, metallurgy, infrared lasers, treatment for unresectable hepatocellular carcinoma, superconductor component

Ho Magnetic flux concentrator, nuclear control rods, solid state lasers, medical lasers for breaking kidney stones

Er Fiber optic cables, fiber amplifiers, fiber lasers, waveguide amplifiers, photographic filter, metallurgy, neutron absorber

Tm Radiation source in portable x‐ray devices, ceramic magnetic devices (ferrites)

Yb Radiation source in portable x‐ray devices, solar cells, metallurgy, fiber lasers, disk lasers

Lu Petroleum cracking catalyst, alkylation, hydrogenation, and polymerization applications, detectors in PET applications

13 consisting of radioactive Sm and a tetraphosphonate chelator used to help relieve bone pain associated with breast, prostate, lung and multiple myeloma cancer.

Gadolinium is unique in having a high neutron absorption cross section coupled with a burn‐up rate that can match that of the 235U isotope used as nuclear fuel. Also, as mentioned before, Gd is used as contrast agent in magnetic resonance imaging due to its unique paramagnetic property with seven electrons in its 4f orbital. Dy has also been used for similar applications, although not as much as Gd.

Rare‐earth doped zeolite (FCC) catalysts are used for “cracking” crude oil to produce shorter‐ chain molecules in petrol refining. Losses of REE through the use of zeolite catalysts in oil refineries might also have contributed to enrichments of REE in atmospheric aerosols during the last century. REE are also used as automotive pollution‐control catalytic converters. Cerium, for example, is a critical component in both gasoline and the newer diesel catalytic converters used in automobiles and trucks. Among other things, the Ce protects the platinum group metals in the converter from oxidation.

The above examples demonstrate the use of REE in a vast range of products and processes. Especially relevant for this thesis are the use of Gd in contrast agents and the production of REE‐ doped zeolites as FCC catalysts. With the wide‐spread usage of REE in various key high‐tech applications, the problem of supply and demand emerges. All of the world’s industrial countries rely almost entirely on import of REE. While high import‐reliance alone is not a risk in the supply chain, the dominance of one country (China) in supplying the raw materials, downstream oxides, associated metals and alloys for the rest of the world is a reason for concern. The situation has worsened given China’s restrictions on REE export due to its internal demand for REE. World REE demand is estimated at 136,000 tons per year, and projected to rise to at least 185,000 tons annually by 2015. With global production estimated around 133,600 tons in 2010 the supply‐ demand difference is covered by previously mined stocks (Humphries, 2011).

As world demand for REE continues to climb (Fig. 1), demand for several specific industries is expected to grow even more. For example, permanent magnet demand is expected to grow by

14 10%‐16% per year over the next several years (e.g. U.S.) due to the growing renewables industry (Humphries, 2011). During the same period, demand for REE in auto catalysts and petroleum cracking catalysts is expected to increase between 6% and 8% each year.

140 Other USA 120 China

100

80

60 REE production (kt/year) production REE 40

20

0 1956 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 Year

Fig. 1. World REE oxide production from 1956 to 2010 (modified after USGS, 2010).

15 5. RARE EARTH ELEMENT TOXICITY

As REE behave coherently in nature due to their trivalent nature and similar ionic radii, their ecotoxicological effects and biological pathways often show similarities. Although toxicological studies have been conducted on single REE, often using a few REE as representatives, the results can be generally true for different REE due to their coherent behavior. For instance, La and Gd are both Ca inhibitors (Weiss and Goodmann, 1975) and affect biological functions that rely on Ca: decrease in the fertility of sea urchins (Oral et al., 2010) or contraction of smooth, skeletal and cardiac muscle, transmission of nerve impulses and blood coagulation (Evans, 1990). Adverse effects for La include a tenfold increase in La concentrations in body tissue of rats after oral administration of La carbonates (Lacour et al., 2005; 2007), accumulation of La in the livers of normal and uremic rats (Slatopolsky et al., 2005), liver damage at La concentrations of 20 mg/kg (Chen et al., 2003) and significant adverse effects on the growth and reproduction of worms (Zhang et al., 2010). Similarly Gd inhibits the reticuloendothelial system, while GdCl3 kills Kupffer cells (Evans, 1990). The combination of Gd‐CA administration at high dosage and renal insufficiency has been linked (Broome, 2008, and references therein) to the disorder known as nephrogenic systemic fibrosis (NSF), a notable example of synergistic effects not foreseen in traditional toxicological tests. First recognized in 1997 (Cowper et al., 2000), NSF patients rose to more than 400 by 2007 (Thakral et al., 2007). Transmetallation with metals within the body is a potential problem (Abraham et al., 2008; Idée et al., 2006), although the results of Möller and Dulski (2010a,b) suggest that this might be a minor problem. Anomalously high concentrations of Gd have also been found in bone tissues, and are reported to persist even 8 years after exposure (Darrah et al., 2009). Sm causes pathological effects in the kidneys and livers of rats (Shi et al., 2006a; 2006b) and LREE were shown to bioaccumulate in carp (Hao et al., 1996).

Despite the large list of studies on REE toxicity, there is hardly any work on the effects of long‐ term exposure at low‐dosage as well as studies of synergistic effects. Without this information there may potentially be unforeseen disorders or ecotoxicological impairment similar to that of NSF. Due to the lack of information about the environmental and toxicological effects of anthropogenic REE, and the prediction that more REE will enter the hydrosphere in the near future as emerging contaminants, the urgency of this study and others with similar research goals is once again emphasized.

16 6. STUDY AREA

The Rhine River is a 1,233 km long river that drains an area of approximately 200,000 km2 over 9 countries. Long‐term average discharge (MQ) ranges from 338 m³/s at the Konstanz gauging station (Lake Constance) to 2,270 m³/s at Rees upstream of the German‐Dutch border (UN, 2007). A population of 58 million inhabitants live within the catchment of the Rhine River, and up to 30 million receive drinking water from the Rhine River or its tributaries (IAWWRB, 2012).

The rivers Aare, Main, Mosel and Neckar are the main tributaries. Together these four rivers host approximately 20 million inhabitants in their catchment, roughly a third of the total population of the Rhine River catchment. As a transnational shippable river, anthropogenic impact on the Rhine River has always been significant.

The Rhine is divided into sub‐catchments as follows. The Rivers Vorderrhein and Hinterrhein form the Alpine Rhine (upstream of Lake Constance), an important contributor to the discharge in the Rhine River due to melt water from the Alps in the summer. The High Rhine is between Lake Constance and Basel (Rhine‐km 0 – 165). The Aare River with its high discharge joins the Rhine River in this section (Rhine‐km 102). The Upper Rhine (Rhine‐km 165 – 529) flows northward from Basel to Bingen along the major rift bearing the same name (Upper Rhine Graben) with the prevalent lithology being alluvial deposits (sandstone) from the Holocene and Pleistocene. The river Neckar and Main converge with the Rhine River in this section, at Rhine‐ km 428 and 497, respectively. Within the next section (Middle Rhine Rhine‐km 529 – 659) from Bingen to Bonn, the Moselle River joins the Rhine River at Rhine‐km 592. This section of the Rhine River shows lithologies of predominantly siliciclastic rocks (shales and claystones). Further downstream, the Lower Rhine (Rhine‐km 659 – 858) is covered by alluvial deposits from the Pleistocene and flows in wide meanders towards the German‐Dutch border.

The Rhine River and its main tributaries have been sampled during three different sampling campaigns in May 2008, May 2009 and October 2009. Fig. 2 shows a map of the study area including sampling locations. The October sampling campaign was during a period of low discharge (688 m3/s at Köln), while discharge was high during May 2008 (1,800 m3/s) and May 2009 (1,670 m3/s).

17 Fig. 2. Map of sampling locations for the Rhine River (blue symbols) and its tributaries (red symbols). Tributaries are labelled in grey text while approximate Rhine‐km for selected locations is given in blue text. Two samples from Lake Constance are also shown (green symbols).

18 7. REFERENCES

Abraham, J.L., Thakral, C., Skov, L., Rossen, K., Marckmann, P., 2008. Dermal inorganic gadolinium concentrations: evidence for in vivo transmetallation and long‐term persistence in nephrogenic systemic fibrosis. Brit J Dermatol 158, 273‐280.

Asher, C.J., 1991. Beneficial elements, functional nutrients and possible new essential elements, in: Mortvedt, J.J., Cox, F.R., Shuman, L.M., Welch, R.M. (Eds.), Micronutrients in agriculture, 2 ed. Soil Science Society of America, Madison, pp. 703–723.

Bau, M., 1996. Controls on the fractionation of isovalent trace elements in magmatic and aqueous systems: evidence from Y/Ho, Zr/Hf, and lanthanide tetrad effect. Contributions to Mineralogy and Petrology V123, 323‐333.

Bau, M., 1999. Scavenging of dissolved yttrium and rare earths by precipitating iron oxyhydroxide: experimental evidence for Ce oxidation, Y‐Ho fractionation, and lanthanide tetrad effect. Geochimica et Cosmochimica Acta 63, 67‐77.

Bau, M., Dulski, P., 1996. Anthropogenic origin of positive gadolinium anomalies in river waters. Earth and Planetary Science Letters 143, 245‐255.

Bau, M., Möller, P., Dulski, P., 1997. Yttrium and lanthanides in eastern Mediterranean seawater and their fractionation during redox‐cycling. Marine Chemistry 56, 123‐131.

Bau, M., Alexander, B., 2006. Preservation of primary REE patterns without Ce anomaly during dolomitization of Mid‐Paleoproterozoic limestone and the potential re‐establishment of marine anoxia immediately after the "Great Oxidation Event". South African Journal of Geology 109, 81‐ 86.

Bau, M., Knappe, A., Dulski, P., 2006. Anthropogenic gadolinium as a micropollutant in river waters in Pennsylvania and in Lake Erie, northeastern United States. Chemie der Erde ‐ Geochemistry 66, 143‐152.

Broome, D.R., 2008. Nephrogenic systemic fibrosis associated with gadolinium based contrast agents: A summary of the medical literature reporting. European Journal of Radiology 66, 230‐ 234.

19 Brown, P.H., Rathjen, A.H., Graham, R.D., Tribe, D.E., 1990. Rare earth elements in biological systems, in: Gschneidner, K.A.J., Eyring, L. (Eds.), Handbook on the physic and chemistry of rare earth. Elsevier Science Publ., Amsterdam, The Netherlands, pp. 423–452.

Buesseler, K.O., Bauer, J.E., Chen, R.F., Eglinton, T.I., Gustafsson, O., Landing, W., Mopper, K., Moran, S.B., Santschi, P.H., VernonClark, R., Wells, M.L., 1996. An intercomparison of cross‐flow filtration techniques used for sampling marine colloids: Overview and organic carbon results. Marine Chemistry 55, 1‐31.

Byrne, R.H., Kim, K.‐H., 1990. Rare earth element scavenging in seawater. Geochimica et Cosmochimica Acta 54, 2645‐2656.

Chang, J., 1991. Effects of lanthanum on the permeability of root plasmalemma and the absorption and accumulation of nutrients in rice and wheat. Plant Physiology Communications 27, 17‐21.

Chen, D., Liu, Y., Chen, A.‐j., Nie, Y.‐X., 2003. Experimental Study of subchronic toxicity of lanthanum nitrate on liver in rats. Nonlinearity Biol Toxicol Med. 1, 469‐480.

Cowper, S.E., Robin, H.S., Steinberg, S.M., Su, L.D., Gupta, S., LeBoit, P.E., 2000. Scleromyxoedema‐like cutaneous diseases in renal‐dialysis patients. Lancet 356, 1000‐1001.

Darrah, T.H., Prutsman‐Pfeiffer, J.J., Poreda, R.J., Campbell, M.E., Hauschka, P.V., Hannigan, R.E., 2009. Incorporation of excess gadolinium into human bone from medical contrast agents. Metallomics 1, 479‐488.

De Baar, H.J.W., Brewer, P.G., Bacon, M.P., 1985. Anomalies in rare earth distributions in seawater: Gd and Tb. Geochimica et Cosmochimica Acta 49, 1961‐1969.

De Baar, H.J.W., Schijf, J., Byrne, R.H., 1991. Solution chemistry of the rare earth elements in seawater. European Journal of Solid State and Inorganic Chemistry 28, 357‐373.

Dupré, B., Viers, J., Dandurand, J.‐L., Polve, M., Bénézeth, P., Vervier, P., Braun, J.‐J., 1999. Major and trace elements associated with colloids in organic‐rich river waters: ultrafiltration of natural and spiked solutions. Chemical Geology 160, 63‐80.

Elbaz‐Poulichet, F., Seidel, J.‐L., Othoniel, C., 2002. Occurrence of an anthropogenic gadolinium anomaly in river and coastal waters of Southern France. Water Res 36, 1102‐1105.

20 Elderfield, H., Upstill‐Goddard, R., Sholkovitz, E.R., 1990. The rare earth elements in rivers, estuaries, and coastal seas and their significance to the composition of ocean waters. Geochimica et Cosmochimica Acta 54, 971‐991.

Evans, C.H., 1990. Biochemistry of the lanthanides. Plenum Press, New York.

Gaillardet, J., Viers, J., Dupré, B., 2003. Trace Elements in River Waters, in: Heinrich, D.H., Karl, K.T. (Eds.), Treatise on Geochemistry. Pergamon, Oxford, pp. 225‐272.

Gerard, M., Seyler, P., Benedetti, M.F., Alves, V.P., Boaventura, G.R., Sondag, F., 2003. Rare earth elements in the Amazon basin. Hydrological Processes 17, 1379‐1392.

Goldstein, S.J., Jacobsen, S.B., 1988. REE in the Great Whale River estuary, northwest Quebec. Earth and Planetary Science Letters 88, 241‐252.

Guo, B., 1988. Rare earth in agriculture. China Agricultural Science and Technology Press, Beijing.

Gursky, A., 1999. Rhein II [Image]. Available online at http://www.popphoto.com/news/2011/11/andreas‐gursky‐photo‐rhein‐ii‐sells‐43‐million‐ auction. Last accessed on May 25, 2012.

Gustafsson, O., Buesseler, K.O., Gschwend, P.M., 1996. On the integrity of cross‐flow filtration for collecting marine organic colloids. Marine Chemistry 55, 93‐111.

Hao, S., Wang, X.R., Hua, Z.Z., Wu, C.H., Wang, L.S., Dai, L.M., Li, Z., Chen, Y.J., 1996. Bioconcentration and elimination of five light rare earth elements in carp (Cyprinus carpio L). Chemosphere 33, 1475‐1483.

Hoffmann, S.R., Shafer, M.M., Babiarz, C.L., Armstrong, D.E., 2000. A critical evaluation of tangential flow ultrafiltration for trace metal studies in freshwater systems. 1. Organic carbon. Environmental Science & Technology 34, 3420‐3427.

Horowitz, A.J., Lum, K.R., Garbarino, J.R., Hall, G.E.M., Lemieux, C., Demas, C.R., 1996. The effect of membrane filtration on dissolved trace element concentrations. Water, Air, & Soil Pollution 90, 281‐294.

Hoyle, J., Elderfield, H., Gledhill, A., Greaves, M., 1984. The behaviour of the rare earth elements during mixing of river and sea waters. Geochimica et Cosmochimica Acta 48, 143‐149.

21 Hugo, V., 1845. The Rhine Volume I. Available online at http://www.archive.org/stream/victorhugosworks037671mbp/victorhugosworks037671mbp_dj vu.txt. Last accessed on May 25, 2012.

Humphries, M., 2011. Rare Earth Elements: The Global Supply Chain. CRS Report for Congress R41347. Washington D.C.: Congressional Research Service; 2011. Available online at www.fas.org/sgp/crs/natsec/R41347.pdf. Last accessed on April 13, 2012.

ICPR. International catchment area of the Rhine: Properties, assessment of environ‐ mental effects of human activities and economic analysis of water use (part A). “Internationaal stroomgebiedsdistrict Rijn: Kenmerken, beoordeling van de milieueffecten van menselijke activiteiten en economische analyse van het watergebruik (deel A)”. Den Haag, the Netherlands: ICBR; 2005. p. 83 . Available online at http://www.iksr.org/fileadmin/user_upload/Dokumente_nl/Rijnkaart/cc_02‐ 05nl_rev._18.03.05_online.pdf

Idée, J.M., Port, M., Raynal, I., Schaefer, M., Le Greneur, S., Corot, C., 2006. Clinical and biological consequences of transmetallation induced by contrast agents for magnetic resonance imaging: a review. Fund Clin Pharmacol 20, 563‐576.

Ingri, J., Widerlund, A., Land, M., Gustafsson, O., Andersson, P., Ohlander, B., 2000. Temporal variations in the fractionation of the rare earth elements in a boreal river; the role of colloidal particles. Chemical Geology 166, 23‐45.

Kim, K.‐H., Byrne, R.H., Lee, J.H., 1991. Gadolinium behavior in seawater: a molecular basis for gadolinium anomalies. Marine Chemistry, 107‐120.

Knappe, A., Möller, P., Dulski, P., Pekdeger, A., 2005. Positive gadolinium anomaly in surface water and ground water of the urban area Berlin, Germany. Chemie der Erde ‐ Geochemistry 65, 167‐189.

Kulaksız, S., Bau, M., 2007. Contrasting behaviour of anthropogenic gadolinium and natural rare earth elements in estuaries and the gadolinium input into the North Sea. Earth and Planetary Science Letters 260, 361‐371.

Kümmerer, K., Helmers, E., 2000. Hospital effluents as a source of gadolinium in the aquatic environment. Environmental Science & Technology 34, 573‐577.

22 Künnemeyer, J., Terborg, L., Meermann, B., Brauckmann, C., Scheffer, I.M.A., Karst, U., 2009. Speciation analysis of gadolinium chelates in hospital effluents and wastewater treatment plant sewage by a Novel HILIC/ICP‐MS method. Environmental Science & Technology 43, 2884‐2890.

Lacour, B., Lucas, A., Auchere, D., Ruellan, N., de Serre Patey, N.M., Drueke, T.B., 2005. Chronic renal failure is associated with increased tissue deposition of lanthanum after 28‐day oral administration. Kidney Int 67, 1062‐1069.

Lacour, B., Nikolov, I.G., Joki, N., Drüeke, T.B., 2007. Lanthanum carbonate, body lanthanum accumulation and potential liver toxicity. Nephrology Dialysis Transplantation 22, 1788‐1789.

Lawrence, M.G., Kamber, B.S., 2006. The behaviour of the rare earth elements during estuarine mixing‐‐revisited. Marine Chemistry 100, 147‐161.

Lawrence, M.G., Ort, C., Keller, J., 2009. Detection of anthropogenic gadolinium in treated wastewater in South East Queensland, Australia. Water Res 43, 3534‐3540.

Lawrence, M.G., 2010. Detection of anthropogenic gadolinium in the Brisbane River plume in Moreton Bay, Queensland, Australia. Marine Pollution Bulletin 60, 1113‐1116.

Lawrence, M.G., Keller, J., Poussade, Y., 2010. Removal of magnetic resonance imaging contrast agents through advanced water treatment plants. Water Sci Technol 61, 685‐692.

Morteani, G., Möller, P., Fuganti, A., Paces, T., 2006. Input and fate of anthropogenic estrogens and gadolinium in surface water and sewage plants in the hydrological basin of Prague (Czech Republic). Environmental Geochemistry and Health 28, 257‐264.

Möller, P., Dulski, P., 2010a. Transmetallation of Gd‐DTPA by Cu, Y and lanthanides and its impact on the hydrosphere. Applied Geochemistry 25, 48‐59.

Möller, P., Dulski, P., 2010b. Gd‐DTPA in the hydrosphere: Kinetics of transmetallation by ions of rare earth elements, Y and Cu. Chem Erde‐Geochem 70, 125‐136.

Möller, P., Dulski, P., Bau, M., Knappe, A., Pekdeger, A., Sommer‐von Jarmersted, C., 2000. Anthropogenic gadolinium as a conservative tracer in hydrology. J Geochem Explor 69‐70, 409‐ 414.

Möller, P., Paces, T., Dulski, P., Morteani, G., 2002. Anthropogenic Gd in surface water, drainage system, and the water supply of the city of Prague, Czech Republic. Environmental Science & Technology 36, 2387‐2394.

23 Möller, P., Morteani, G., Dulski, P., 2003. Anomalous gadolinium, cerium, and yttrium contents in the Adige and Isarco River waters and in the water of their tributaries (Provinces Trento and Bolzano/Bozen, NE Italy). Acta hydrochimica et hydrobiologica 31, 225‐239.

Möller, P., Knappe, A., Dulski, P., Pekdeger, A., 2011. Behavior of Gd‐DTPA in simulated bank filtration. Applied Geochemistry 26, 140‐149.

Nozaki, Y., Lerche, D., Alibo, D.S., Snidvongs, A., 2000. The estuarine geochemistry of rare earth elements and indium in the Chao Phraya River, Thailand. Geochimica et Cosmochimica Acta 64, 3983‐3994.

Oral, R., Bustamante, P., Warnau, M., D'Ambra, A., Guida, M., Pagano, G., 2010. Cytogenetic and developmental toxicity of cerium and lanthanum to sea urchin embryos. Chemosphere 81, 194‐ 198.

Petelet‐Giraud, E., Klaver, G., Negrel, P., 2009. Natural versus anthropogenic sources in the surface‐ and groundwater dissolved load of the Dommel river (Meuse basin): Constraints by boron and strontium isotopes and gadolinium anomaly. J Hydrol 369, 336‐349.

Pokrovsky, O.S., Schott, J., 2002. Iron colloids/organic matter associated transport of major and trace elements in small boreal rivers and their estuaries (NW Russia). Chemical Geology 190, 141‐179.

Rabiet, M., Brissaud, F., Seidel, J.‐L., Pistre, S., Elbaz‐Poulichet, F., 2005. Deciphering the presence of wastewater in a medium‐sized Mediterranean catchment using a multitracer approach. Applied Geochemistry 20, 1587‐1596.

Rabiet, M., Togola, A., Brissaud, F., Seidel, J.L., Budzinski, H., Elbaz‐Poulichet, F., 2006. Consequences of treated water recycling as regards pharmaceuticals and drugs in surface and ground waters of a medium‐sized Mediterranean catchment. Environmental Science & Technology 40, 5282‐5288.

Rabiet, M., Brissaud, F., Seidel, J.L., Pistre, S., Elbaz‐Poulichet, F., 2009. Positive gadolinium anomalies in wastewater treatment plant effluents and aquatic environment in the Herault watershed (South France). Chemosphere 75, 1057‐1064.

Redling, K., 2006. Rare Earth Elements in Agriculture with Emphasis on Animal Husbandry, Faculty of Veterinary Medicine. LMU Muenchen, Muenchen, p. 326.

24 Shiller, A.M., 2002. Seasonality of dissolved rare earth elements in the lower Mississippi River. Geochem Geophy Geosy 3.

Sholkovitz, E.R., 1978. The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co and Cd during estuarine mixing. Earth and Planetary Science Letters 41, 77‐86.

Sholkovitz, E.R., Boyle, E.A., Price, N.B., 1978. The removal of dissolved humic acids and iron during estuarine mixing. Earth and Planetary Science Letters 40, 130‐136.

Sholkovitz, E.R., 1992. Chemical evolution of rare earth elements: fractionation between colloidal and solution phases of filtered river water. Earth and Planetary Science Letters 114, 77‐ 84.

Sholkovitz, E.R., 1993. The geochemistry of rare earth elements in the Amazon River estuary. Geochimica et Cosmochimica Acta 57, 2181‐2190.

Sholkovitz, E.R., 1995. The aquatic chemistry of rare earth elements in rivers and estuaries. Aquatic Geochemistry V1, 1‐34.

Sholkovitz, E., Szymczak, R., 2000. The estuarine chemistry of rare earth elements: comparison of the Amazon, Fly, Sepik and the Gulf of Papua systems. Earth and Planetary Science Letters 179, 299‐309.

Shi, W.L., Shen, X.Y., Ma, X.Y., 2006a. Effects of samarium on liver and kidney of rats. J Rare Earth 24, 415‐418.

Shi, W.L., Shen, X.Y., Ma, X.Y., 2006b. Influence of samarium on learning and memory function of rats. J Rare Earth 24, 419‐422.

Slatopolsky, E., Liapis, H., Finch, J., 2005. Progressive accumulation of lanthanum in the liver of normal and uremic rats. Kidney Int 68, 2809‐2813.

Thakral, C., Alhariri, J., Abraham, J.L., 2007. Long‐term retention of gadolinium in tissues from nephrogenic systemic fibrosis patient after multiple gadolinium‐enhanced MRI scans: case report and implications. Contrast Media Mol I 2, 199‐205.

Tricca, A., Stille, P., Steinmann, M., Kiefel, B., Samuel, J., Eikenberg, J., 1999. Rare earth elements and Sr and Nd isotopic compositions of dissolved and suspended loads from small river systems in the Vosges mountains (France), the river Rhine and groundwater. Chemical Geology 160, 139‐ 158.

25 Verplanck, P.L., Taylor, H.E., Nordstrom, D.K., Barber, L.B., 2005. Aqueous stability of gadolinium in surface waters receiving sewage treatment plant effluent, Boulder Creek, Colorado. Environmental Science & Technology 39, 6923‐6929.

Verplanck, P.L., Furlong, E.T., Gray, J.L., Phillips, P.J., Wolf, R.E., Esposito, K., 2010. Evaluating the Behavior of Gadolinium and Other Rare Earth Elements through Large Metropolitan Sewage Treatment Plants. Environmental Science & Technology 44, 3876‐3882.

Wang, Z., Liu, D., Lu, P., Wang, C., 2001. Accumulation of Rare Earth Elements in Corn after Agricultural Application. Journal of Environmental Quality 30, 37‐45.

Weiss, G.B., Goodman, F.R., 1975. Interactions between Several Rare‐Earth Ions and Calcium‐Ion in Vascular Smooth‐Muscle. J Pharmacol Exp Ther 195, 557‐564.

Wu, Z., X.Tang, Tsui, C., 1983. Studies on the effect of rare earth elements on the increasement of yield in agriculture. Journal of the Chinese Rare Earth Society 1, 70‐75.

Wu, Z., Li, J., Xu, J., Xin, S., 1985. The effect of nodulation and nitrogen fixation of soybean plant, International Conference on Rare Earth Development and Applications. China Science Press, Beijing, pp. 1505‐1511.

Xiong, B., 1995. Application of the rare earths in Chinese agriculture and their perspective development, Rare Earths in Agriculture. Australian Academy of Technological Sciences and Engineering, Victoria, Australia, pp. 5–9.

Zhang, H., He, X., Bai, W., Guo, X., Zhang, Z., Chai, Z., Zhao, Y., 2010. Ecotoxicological assessment of lanthanum with Caenorhabditis elegans in liquid medium. Metallomics 2, 806‐810.

Zhang, J., Nozaki, Y., 1998. Behavior of rare earth elements in seawater at the ocean margin: a study along the slopes of the Sagami and Nankai troughs near Japan. Geochimica et Cosmochimica Acta 62, 1307‐1317.

Zhu, Y., Hoshino, M., Yamada, H., Itoh, A., Haraguchi, H., 2004. Gadolinium anomaly in the distributions of rare earth elements observed for coastal seawater and river waters around Nagoya City. Bulletin of the Chemical Society of Japan 77, 1835‐1842.

Zhu, Y., Hattori, R., Rahmi, D., Itoh, S.O., Fujimori, E., Umemura, T., Haraguchi, H., 2005. Fractional distributions of trace metals in surface water of Lake Biwa as studied by ultrafiltration and ICP‐MS. Bulletin of the Chemical Society of Japan 78, 1970‐1976.

26 Online Resources:

First Assessment of Transboundary Rivers, Lakes and Groundwaters (UN, 2007) http://www.unece.org/fileadmin/DAM/env/water/blanks/assessment/assessmentweb_full.pdf

International Association of Water Works in the Rhine Basin, 2012: http://www.iawr.org

National Industrial Notification and Assessment Scheme, 2011: http://www.nicnas.gov.au/publications/car/new/na/nafullr/na0800fr/na899fr.pdf

Phoslock Water Solutions Limited, 2011: http://www.phoslock.com.au/irm/content/annualreport/full.pdf

USGS, 2010. Global Rare Earth Oxide (REO) Production Trends: http://minerals.usgs.gov/minerals/pubs/commodity/rare_earths/ree‐trends‐2010.pdf

27 ACKNOWLEDGEMENTS

I am grateful to my supervisor Michael Bau, for his continuous support throughout the last decade  and providing me with challenging scientific endeavors. Also for his invaluable know‐ how related to geochemistry as well as obscure psychedelic bands from the 60s and 70s. I enjoyed the many excursions and sampling campaigns that much more thanks to him.

Special thanks go to Andrea Koschinsky for her energetic and positive presence in our working group and Dieter Garbe‐Schönberg for making time for my thesis on short notice.

I would like to recognize the support of Jacobs University Bremen, financial and otherwise. 11 years is a long time…

This project was funded by the German Science Foundation (DeutscheForschungsGemeinschaft) Grant no BA 2289/2‐1.

Deutscher Akademischer Austauschdienst (DAAD) provided support for academic exchange to Santa Maria, Rio Grande do Sul (Brazil).

I also thank the following people:

Gila Merschel for her help and company during the sampling campaign of October 2009.

Jule Mawick, Daniela Meissner and Annika Moje for their constant help in the Geochemistry laboratory and all OceanLab staff for being part of the extended family.

Brian Alexander, for always having time to answer questions, his invaluable help with the ICP‐ MS, and for being a good drinking buddy.

Past and present PhDs in our working group, Katja Schmidt, Geerd Smidt, Nathalie Tepe, Sebastian Viemann and Adilah Ponnurangam for their ideas and assistance. Especially Geerd, with whom we shared a little more on several occasions 

Finally Agata, for her constant support, patience and being there when the going got tough.

28 This thesis represents original and independently conducted research that has not been submitted to any other university for the conferral of a degree.

Serkan Kulaksız

Bremen, Germany – 20th May 2012

Resubmitted with changes on 5th June 2013

29 30 CHAPTER II – SAMPLING AND METHODS

1. SAMPLE PREPARATION

Each river water sample is taken in a 1 L acid‐cleaned polyethylene (PE) bottle. Each bottle is rinsed with several sample volumes before the final sample is taken. The sample is stored in a dark and cool container until brought into the laboratory. Once in the laboratory, each sample is filtered through a 0.2 µm filter (mixed cellulose ester, Schleicher & Schuell) inside a Sartorius filtration unit. The filtration unit and the filter are pre‐cleaned with dilute HCl (0.01 M, ultrapure) followed by copious amounts of deionized (DI, ultrapure) water. The first 100 mL filtration volume of each sample serves to equilibrate the filtration unit with the sample and is then

‐ discarded. The next 100 mL of filtered sample volume is used for bicarbonate (HCO3 ) titration. The remaining sample volume is filtered before an aliquot of 10 mL is taken for major anion

Sample

HCO titration first 100 mL Filtration 3 discarded (0.2 µm) (100 mL)

ICP‐OES Acidification IC Analysis 10 mL (pH 1.8‐2.0) (10 mL)

“Original” Spike addition Pre‐concentration (C cartridge) 10 mL (Tm) 18

Internal Standard addition “Eluate” (Rb, Ru, Bi)

Internal Standard addition Evaporation ICP‐MS (Rb, Ru, Bi) Uptake in HCl

Fig. 1. Flow chart showing the sample preparation procedure

31 analysis using ion chromatography (IC). The sample is then acidified to pH 1.8‐2.0 using a few mL of 6 M HCl (ultrapure) before a 10 mL aliquot is taken for analysis of major cations using inductively coupled plasma optical emission spectrometry (ICP‐OES). Thulium (Tm, 1 mL 250 ppb) is added to the remaining sample for quantification of sample recovery. Another aliquot of 10 mL is taken for analysis of non‐REY elements and Tm using ICP‐MS before the remaining sample volume is passed through an ion‐exchange column for REY pre‐concentration and matrix separation. The 20 mL aliquot is labeled “original” while the volume that is passed through the ion‐exchange column is labeled “eluate”. The eluate is evaporated to incipient dryness before being taken up in 6M HCl. For an overview of the sample preparation please see Fig. 1.

Pre‐concentration of REY is necessary because reliable quantification of REY without pre‐ concentration poses a challenge due to the low concentrations of REY in filtered river waters. Also the pre‐concentration method removes all matrix elements from the samples and minimizes interference during ICP‐MS analysis. Most notably, this method removes Ba, an element that creates polyatomic (BaO and to a lesser extent BaOH) interference on Eu. Without the removal of Ba from the matrix, Eu determination would not be possible. Additionally, corrections are applied for polyatomic interference of BaCl on Tm, Yb and Lu. Relative interference correction factors for all REE are below 5%, except for Yb and Lu which are occasionally higher (below 20% and 10% respectively).

1.1 PRE‐CONCENTRATION

The method used for pre‐concentration in the geochemistry laboratory at Jacobs University (GeoChemLab) is adapted from Bau and Dulski (1996) and is based on the original method of Shabani et al. (1992). The acidified (pH 1.8‐2.0) main sample volume with Tm spike is passed through a C18 cartridge (Waters, Sep‐Pak® Classic C18, single use) freshly prepared as follows:

Each cartridge is cleaned with 10 mL of 6M HCl (ultrapure) using a peristaltic pump (Masterflex) running at 3 mL/min using tubing pre‐cleaned with 6M HCl (ultrapure) and DI water. To remove the acid, 10 mL of water is passed through the cartridge at 3 mL/min. Next, 0.375 mL of phosphate ester (Merck®, 2‐ethylhexyl phosphate) is passed through the cartridge in reverse order at ~ 8 mL /min. The cartridge is cleaned with 10 mL of 6M HCl (ultrapure) at 3 mL/min

32 before it is neutralized with 40 mL of DI at approximately 10 mL/min. The speed with which DI is passed over the cartridge in this final step must not exceed the limit at which water starts flowing without single drops forming at the end of the cartridge.

The prepared C18 cartridge quantitatively retains REY due to sorption onto the phosphate ester it is loaded with. After each sample is passed over the column at 15 mL/min, sample waste is collected for weighing the amount of sample. In order to remove any remaining matrix elements, 10 mL of 0.01 M HCl (ultrapure) is passed over the column at 3 mL/min. The dilute acid is strong enough to remove remaining mono‐ and divalent matrix elements from the ester‐loaded cartridge. Finally, quantitative removal of REY is achieved by passing 40 mL of 6M HCl (ultrapure) over the cartridge at 3 mL/min (eluate).

The eluate is then evaporated in pre‐cleaned Teflon® beakers at 100 oC to incipient dryness. The sample residue is dissolved in 1mL 6M HCl (ultrapure) and transferred to a sample vessel.

1.1 1.2 1.3 0.375 mL 1.4 1.5 10 mL 10 mL 10 mL 40 mL Phosphate ester 6 M HCl DI 6 M HCl DI ~ 8 mL/min 3 mL/min 3 mL/min 3 mL/min ~ 10 mL/min (reverse direction)

2 Sample 4 C 40 mL 18 6 M HCl cartridge 3 mL/min

Liquid waste (weighed for sample mass) “Eluate”

Discard (matrix 3 10 mL elements) 0.01 M HCl 3 mL/min

Fig. 2. Flow chart showing the steps necessary for preparing the ion‐exchange column (steps 1.1‐1.5). Step 2 shows the loading of the sample onto the cartridge while step 3 removes matrix elements. Step 4 uses 6M HCl as a solvent to dissolve the REY to form the “eluate” sample volume.

33 Depending on the TDS in the sample material (visible approaching incipient dryness), small volumes of 0.5M HCl (ultrapure) or 6M HCl (ultrapure) are added to the Teflon® beaker and evaporated to incipient dryness before being transferred into the sample vessel. This step is repeated to make sure no sample material is left in the beaker. An internal standard of Ru, Re and Bi (1ppb: 100 µL of 100 ppb in 10 mL) is added to the sample before dilution to a total volume of 10 mL in 0.5 M HCl matrix. Similarly, the “original” (10 mL filtered, acidified, spiked) sample aliquot is diluted 1:10 to a 10 mL sample in a 0.5 M HCl matrix, containing the internal standard (Ru, Re, Bi) at Ru, Re and Bi concentrations of 1 ppb (100 µL of 100 ppb in 10 mL). Both the eluate and original sample must be analyzed immediately after dilution in the sample vessel, due to the possibility of loss of REY to the vessel walls. For an overview of the preparation and application of the ion‐exchange column, see Fig. 2. The evaporation step provides increased pre‐ concentration from approximately 20 fold (only ion‐exchange) to approximately 80‐fold (ion‐ exchange + evaporation) and allows the dilution of the acid matrix from 6 M HCl to 0.5 M HCl. Since quantification of REY using ICP‐MS is most effective with samples in dilute acid matrices, such as 0.5M HCl or 0.5M HNO3, this step is necessary.

2. METHODS

2.1 HCO3 TITRATION

Water hardness and HCO3‐ concentration is determined by titrating 100 mL of filtered unacidified sample volume with 0.1 M HCl to an endpoint of pH 4.3. The amount of 0.1 M HCl (in mL) used for the titration to the endpoint (pH 4.3) is the concentration of HCO3 in mval/L.

2.2 ION CHROMATOGRAPHY

A compact ion chromatograph (IC) from Metrohm (761 Compact IC) is used for measuring F‐, Cl‐,

‐ ‐ 2‐ Br , NO3 and SO4 in natural river waters. The IC is equipped with a separating column (Metrosup A Supp 5 250) and a suppressor module (MSM) for chemical suppression to record every new chromatogram under comparable conditions. The suppressor consists of three suppressor units (suppression, regeneration with sulfuric acid and rinsing with water) and is automatically triggered before each sample is measured. A mixture of 3.2 mM Na2CO3 and 1 mM NaHCO3 is used as the eluent, while H2O and H2SO4 (20 mM) are used as the rinse solution and the regeneration solution. Samples are diluted 1:10 before analysis with the IC. Due to the high

34 number of samples processed, not all samples could be measured at the same time. Several samples are measured twice (once a few days after sampling, once more after the last sample is measured) in order to make sure no elements were lost in the filtered and unacidified 10 mL sample volumes.

2.3 INDUCTIVELY COUPLED PLASMA – OPTICAL EMISSION SPECTROMETER

GeoChemLab is equipped with an ICP‐OES (Spectro CIROS SOP) for routine measurements of

2+ + + 2+ Ca , K , Na , Mg , Al, Cu, Fe, Mn, P, Si and Zn. In the natural river water samples presented here, Ca2+, K+, Mg2+, Na+ and Si were reliably measured while Cu, Fe, Mn, P and Zn were consistently below the limit of quantification. The filtered and acidified samples were diluted 1:10 before analysis using ICP‐OES.

2.4 INDUCTIVELY COUPLED PLASMA – MASS SPECTROMETER

GeoChemLab is equipped with an inductively coupled plasma mass spectrometer (ICP‐MS ‐ Perkin‐Elmer/Sciex ELAN DRC‐e). The instrument is routinely used for successful determination of the trace elements Rb, Sr, Y, Ba, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu and U in a variety of water types, including natural river waters. There are several advantages of using an ICP‐MS: The instrument sensitivity is high with most elements in the ng/kg (ppt) range regardless of sample matrix. The liquid sample introduction system is automated and facilitates multi‐ element geochemical analyses of a large number of samples in a short period of time.

Several corrections and quality checks are made during data reduction on raw data from the ICP‐ MS: Firstly, internal standard correction is applied following the method of Doherty (1989) using three elements (Ru, Re and Bi, 1 ppb) to correct for changes due to matrix effects and instrument drift. Secondly, corrections for interfering polyatomic species are made, based on the method of Dulski (1994) for elements with element numbers 151 (Eu) to 183 (W). Thirdly, blank correction is applied based on a method blank treated as any other sample (i.e. all previous corrections also applied). Additionally, in the case of polyisotopic elements, if there are no changes in their natural abundance in nature, concentrations based on different isotopes are used for checking the quality of data. If the concentrations based on single isotopes for a given element differ by

35 more than a user‐defined threshold (5 %), data is flagged and in some cases not included in this dataset.

For a complete list of corrections for interfering polyatomic species, details on the internal and external standards as well as the standard run‐list for the ICP‐MS, see Alexander, 2008 (Table 2, Fig. 2 and Fig. 3 in Technical Report 18).

2.4.1 Method Reliability

10 6

1 REE/ PAAS x 10

Wiembach Creek: 24 Apr 05 0.1 17 Jan 06 20 May 07 29 May 08 08 Sep 08 29 Jan 09 06 May 09 26 Oct 09 (High colloid sample) 0.01 19 Jan 12

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 3. REE plots of Wiembach Creek between 2005 and 2012. Wiembach Creek is a pristine creek and is used as control in the GeochemLab. Each sample is sampled, prepared and analyzed within the same batch of samples processed and reported here. The sample from 2012 has been sampled after heavy rain and presented extremely high amounts of particulate and/or colloidal matter.

Fig. 3 shows Wiembach Creek, a small tributary to the Wupper River in Opladen (Leverkusen), Germany, and serves as the control sample for REE analyses. The creek is pristine and does not

36 A 6 1 REE / PAAS x x 10 / PAAS REE

0.1 Rhine River May 2008 S09-R-67 S09-R-68 S09-R-69

B 6

1 REE / PAAS x 10

Rhine River October 2008 0.1 S09-R-158 S09-R-159 S09-R-160

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Y Ho Er Tm Yb Lu

Fig. 4. Triplicate Rhine River water samples from May 2008 (A) and October 2008 (B). Each triplicate is individually sampled, processed and analyzed. The area shaded is the standard deviation of the three samples. receive any input of anthropogenic REE making it ideal as a long‐term control. In contrast to the majority of samples presented in the subsequent chapters, these samples show no anthropogenic REE, confirming the suitability and reliability of the method to distinguish the presence of even relatively small amounts of anthropogenic REE. The general shape of the 8 REE patterns is

37 Table 1 – REE and Y concentrations (ng/kg) for two sets of duplicates of Rhine River water at Rhine‐ km 823.3

S09‐R‐67 S09‐R‐68 S09‐R‐69 % st dev S09‐R‐158 S09‐R‐159 S09‐R‐160 % st dev La 21.81 20.67 23.56 7 38.93 43.23 41.29 5 Ce 3.86 3.34 3.75 7 4.47 4.91 4.34 6 Pr 0.89 0.84 0.89 3 1.01 1.11 1.14 6 Nd 3.97 3.71 4.18 6 4.53 4.98 5.08 6 Sm 1.11 1.12 1.17 3 1.28 1.31 1.33 2 Eu 0.29 0.29 0.28 1 0.33 0.36 0.33 5 Gd 12.48 13.06 12.42 3 27.87 25.38 26.92 5 Tb 0.25 0.25 0.25 1 0.31 0.29 0.29 5 Dy 1.72 1.74 1.73 1 2.00 1.91 1.97 2 Y 12.73 14.02 13.16 5 19.51 19.90 19.82 1 Ho 0.45 0.44 0.43 2 0.50 0.52 0.54 4 Er 1.54 1.53 1.51 1 1.89 1.86 1.83 2 Tm 0.28 0.28 0.28 1 0.31 0.30 0.31 1 Yb 2.50 2.57 2.49 2 2.65 2.50 2.54 3 Lu 0.46 0.50 0.56 10 0.60 0.54 0.53 7

Samples S09‐R‐67, S09‐R‐68, S09‐R‐69 form a set of duplicates taken in May 2008. Samples S09‐R‐158, S09‐R‐ 159 and S09‐R‐160 form a second set of duplicates (October 2009). All samples were taken at the same location

Table 2 – Certified reference material (SLRS‐4) concentrations and published data

% deviation from published Lawrence & GeoChemLab % st dev data Kamber, 2007

Nov 2007 May 2008 Nov 2007 May 2008 published data La 266.8 267.7 0 9 8 292.0 Ce 372.1 362.3 2 2 1 364.8 Pr 69.0 66.7 2 4 7 71.8 Nd 272.0 265.3 2 1 1 269.0 Sm 56.5 54.8 2 2 5 57.9 Eu 8.0 7.6 4 1 4 7.9 Gd 33.6 33.9 1 3 2 34.8 Tb 4.4 4.4 0 0 0 4.4 Dy 23.8 23.2 2 1 1 23.5 Y 141.4 130.4 6 3 5 137.2 Ho 4.6 4.5 2 5 7 4.8 Er 13.2 12.9 2 2 1 13.0 Tm ‐ ‐ ‐ ‐ ‐ 1.9 Yb 11.5 11.6 1 6 5 12.2 Lu 1.8 1.8 0 6 6 1.9

Two samples prepared and analyzed at the GeoChemLab compared to published data for SLRS‐4.

38 consistent within samples: gradual increase in normalized concentrations from LREE to HREE and a small jump from Gd to Tb. The differences in absolute concentrations are most likely due to the varying contribution of colloidal REE in each sample. While showing the same general features, the data for “19 Jan 2012” shows much higher normalized REE concentrations as compared to all other Wiembach Creek samples. This sample had unusually high turbidity.

Fig. 4 shows individually sampled, processed and analyzed triplicates of Rhine River water from two separate sampling campaigns. There is excellent agreement between the patterns for both sets of triplicates and REE concentrations are in good agreement between the triplicates, with most REE well within the 5% analytical uncertainty. There seems to be slightly higher deviation between duplicates for LREE, especially in Fig. 4b. Except for La, Ce, Pr and Nd (SD < 7%) and Lu (SD < 10%) all other elements show less than 5% deviation between duplicates, in both sets of samples. It must be emphasized that this is not the precision of the ICP‐MS, which is much lower than that (< 1%), but the overall reproducibility of the sample preparation and analysis. Please refer to Table 1 for concentrations and standard deviation of each set of duplicates.

Finally, results of preparation and analysis of the certified reference material (SLRS‐4) for trace elements in river water are shown in Fig. 5. There are no certified concentrations for REE and Y in SLRS‐4, although REY data for this standard has been published (Lawrence and Kamber, 2007). Published data, together with two analyses from the GeoChemLab show excellent agreement. Except for Y (< 6%), standard deviation for all elements between the two GeoChemLab analyses is less than 5%. Most elements are in agreement within 5% of published data (n = 25, Lawrence and Kamber, 2007), while La (< 9%), Pr (< 7%), Ho (< 7%), Yb (< 6%) and Lu (< 6%) show slightly higher deviation. Considering the precision of the duplicates and the general analytical uncertainty (5%), preparation and analyses of SLRS‐4 are in good agreement with published data.

39 10 6

SLRS-4 GeoChem Lab: REE / PAAS x 10 November 2007 May 2008

Lawrence & Kamber, 2007

1 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Y Ho Er Tm Yb Lu

Fig. 5. Two analyses of the river water certified reference material for trace elements (SLRS‐4). The shaded error region is color coded to match the symbol of the analysis it refers to. Published values (Lawrence & Kamber, 2007) are shown for comparison.

40 3. REFERENCES

Alexander, B. W., 2008. Trace element analyses in geological materials using low resolution inductively coupled plasma mass spectrometry (ICPMS). Internal Technical Report 18, Jacobs University.

Bau, M., Dulski, P., 1996. Anthropogenic origin of positive gadolinium anomalies in river waters. Earth and Planetary Science Letters 143, 245‐255.

Dulski, P., 1994. Interferences of oxide, hydroxide and chloride analyte species in the determination of rare earth elements in geological samples by inductively coupled plasma‐mass spectrometry. Fresenius' Journal of Analytical Chemistry 350, 194‐203.

Shabani, M.B., Akagi, T., Masuda, A., 1992. Preconcentration of trace rare‐earth elements in seawater by complexation with bis(2‐ethylhexyl) hydrogen phosphate and 2‐ethylhexyl dihydrogen phosphate adsorbed on a C18 cartridge and determination by inductively coupled plasma mass spectrometry. Analytical Chemistry 64, 737‐743.

41

42 CHAPTER III – RARE EARTH ELEMENTS IN THE RHINE RIVER,

GERMANY: FIRST CASE OF ANTHROPOGENIC LANTHANUM AS A

DISSOLVED MICROCONTAMINANT IN THE HYDROSPHERE

This chapter was published in Vol. 37 of the journal Environment International in 2011 and can be found online at http://dx.doi.org/10.1016/j.envint.2011.02.018.

43 Environment International 37 (2011) 973–979

Contents lists available at ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Short communication Rare earth elements in the Rhine River, Germany: First case of anthropogenic lanthanum as a dissolved microcontaminant in the hydrosphere

Serkan Kulaksız ⁎, Michael Bau

Earth and Space Science Program, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany Integrated Environmental Studies Program, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany article info abstract

Article history: The distribution of dissolved rare earth elements (REE) in the Rhine River, Germany, shows the anthropogenic Received 14 December 2010 gadolinium (Gd) microcontamination that is commonly observed in rivers in densely populated countries Accepted 28 February 2011 with a highly evolved health care system. However, the Rhine River also carries anomalously high Available online 2 April 2011 concentrations of lanthanum (La), which produce very large positive La anomalies in normalized REE distribution patterns. These positive La anomalies first occur north of the City of Worms and then decrease in Keywords: size downstream, but are still significant approximately 400 km downstream, close to the German–Dutch Rare earth elements fl Anthropogenic lanthanum border. The strong La enrichment is of anthropogenic origin and can be traced back to ef uent from a Rhine River production plant for fluid catalytic cracking catalysts at Rhine river-km 447.4. This effluent is characterized by FCC catalyst extremely high dissolved total REE and La concentrations of up to 52 mg/kg and 49 mg/kg, respectively. Such La concentrations are well-above those at which ecotoxicological effects have been observed. The Rhine River is the first case observed to date, where a river's dissolved REE inventory is affected and even dominated by anthropogenic La. Our results suggest that almost 1.5 t of anthropogenic dissolved La is exported via the Rhine River into the North Sea per year. This reveals that the growing industrial use of REE (and other formerly “exotic” elements) results in their increasing release into the environment, and highlights the urgent need to determine their geogenic background concentrations in terrestrial surface waters. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction europium (Eu) are common (Fig. 1A), whereas the occurrence of small positive La and gadolinium (Gd) anomalies is confined to The elements lanthanum (La; Z=57) to lutetium (Lu; Z=71), and seawater (Byrne and Kim, 1990; Kim et al., 1991; de Baar et al., 1991). sometimes including yttrium (Z=39), are usually referred to as the However, in the mid-1990s, anomalously high concentrations of rare earth elements (REE). They are of great economic interest and Gd were detected in several rivers and streams in Germany (Fig. 1B) concern, as they are of crucial importance for numerous high- and, particularly, in the low-discharge Havel River, downstream of the technology products and processes, while world supply is controlled City of Berlin (Bau and Dulski, 1996). Reported Gd concentrations by a monopoly (U.S. GAO, 2010). However, the REE are also widely were up to three orders of magnitude higher than geogenic used as valuable tools in geochemical research. In surface waters they background values, producing a large positive anthropogenic Gd are characterized by coherent behavior, because (i) they are anomaly in the REE pattern of Havel River water. This finding marked exclusively trivalent in low-temperature systems (with the exception the beginning of a new era of REE geochemistry, as the potential of cerium, Ce), and (ii) the ionic radii of the trivalent REE decrease influence of anthropogenic activity on the REE distribution in natural systematically with increasing atomic number. For convenience, their waters expanded REE geochemistry into the field of environmental distribution in natural materials is usually illustrated by normalized aqueous geochemistry. The positive Gd anomalies reported for river REE patterns (REE concentrations in surface waters are normalized to waters worldwide (Fig. 1B) are caused by Gd-based contrast agents Post-Archean Australian Shale, PAAS, McLennan, 1989), in which an used in magnetic resonance imaging (MRI). Due to the high toxicity of anomalous concentration of an individual REE is easily recognized as a aqueous “free” Gd3+, several Gd-based water-soluble contrast agents positive or negative anomaly in an otherwise smooth pattern. In (Gd-CA) have been developed, all of which bind the Gd3+ ion in a pristine river waters that are unaffected by anthropogenic REE input, highly stable organic complex (e.g. with DTPA), ensuring its close-to- only negative anomalies of Ce and, occasionally, anomalies of quantitative excretion from the body. Since the approval of the first Gd-CA (Magnevist®) by the FDA (U.S. Food and Drug Administration) in 1988, more than 100 million Gd contrast enhanced MRI applications ⁎ Corresponding author. Tel.: +49 421 2003228. have been undertaken worldwide (Pering, 2009), and as a consequence E-mail address: [email protected] (S. Kulaksız). anthropogenic Gd anomalies have now been reported from rivers and

0160-4120/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2011.02.018

44 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979

2 Wiembach Creek Rhine River 1 2 Amazon River Havel River 1 Connecticut River Ara River 5 10 1 10 Xijiang River Lake Eerie 3 2 3 Toshibetsu River Ohio River 3 6 Delaware River Elbe River 4 6 Indus RIver Ems River 6 7 Danube River 1 1 REE / PAAS x 10

0.1 0.1

A B 0.01 0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 1. REESN patterns of pristine rivers (A) showing smooth REE patterns without any anthropogenic anomalies and REESN patterns of rivers showing large positive anthropogenic Gd anomalies (B). 1Gaillardet et al., 2003; 2Bau and Dulski, 1996; 3Bau et al., 2006; 4Jacobsen and Goldstein, 1988; 5Nozaki et al., 2000; 6Kulaksız and Bau, 2007, in press 7. Note that in the mid-1990s the Rhine River (Bau and Dulski, 1996) did not show any La anomaly. lakes (Tricca et al., 1999; Nozaki et al., 2000; Elbaz-Poulichet et al., 2002; Zhu et al., 2004; Kulaksız and Bau, 2007; Lawrence, 2010), groundwater Möller et al., 2002, 2003; Knappe et al., 2005; Rabiet et al., 2005, 2009; Zhu (Möller et al., 2000; Knappe et al., 2005;Rabiet et al., 2009)andtapwater et al., 2004, 2005, Verplanck et al., 2005; Bau et al., 2006; KulaksızandBau, (Bau and Dulski, 1996; Möller et al., 2002; Kulaksız and Bau, in press). 2007, in press; Lawrence et al., 2006; Morteani et al., 2006; Petelet-Giraud Hence, while pristine rivers exhibit smooth REE patterns without positive et al., 2009), estuaries and coastal waters (Elbaz-Poulichet et al., 2002; anthropogenic anomalies (Fig. 1A), it is now well established that rivers

6°E 7°E 8°E 9°E 10°E

Xanten 0 50 100 010.5 Lippe 16 km km NL Essen Dortmund Duisburg A Worms-448.8 B Ruhr 42 Düsseldorf Neuss Worms-downstr.WWTP 29 Wupper 41 Wiembach Creek Leverkusen Worms-upstr.WWTP 51°N Köln 1.1 25 Sieg Lahn 33 Bonn 29 Erft Effluent-447.4 B Pfrimm Ahr 1.4 Worms-446.7 Lahnstein 1.3 39 Frankfurt am Main 50°N Nahe Mainz Main Rhine River L 46 Worms Pfrimm Creek D Mannheim 1 Mosel 49°N

Stuttgart Worms Strasbourg ! Neckar F Strasbourg 1.5 Ill 48°N Sample Site Lake Constance ! City CH A

Fig. 2. Map showing sampling locations along the Rhine River (A) and detailed map of the area in the vicinity of a FCC catalyst plant at Rhine river-km 447.4 (B). Numbers below sample sites show the LaSN/LaSN* ratio of the respective sample to quantify the size of the anthropogenic positive La anomaly.

45 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979 within densely populated catchments in countries with highly developed Constance) were collected at different times between January 2009 healthcare systems (Bau et al., 2006) exhibit large positive anthropogenic and October 2009. The latter group of samples was taken from various Gd anomalies (Fig. 1B). sites within effluent plumes originating from (i) a waste water The Rhine River in Germany is no exception to this and shows treatment plant (WWTP) at river-km 448.4, that serves Worms, and significant positive Gd anomalies. However, we report here the first (ii) a discharge pipe of a company located at river-km 447.4 (samples observation of yet another type of anthropogenic aqueous REE EF1 to EF7), which produces fluid catalytic cracking (FCC) catalysts. In microcontamination in surface water: downstream from the City of January, May and October of 2009, several samples of this effluent Worms (Fig. 2A), the Rhine River shows dissolved La concentrations were taken as close to the discharge pipe as possible in order to that are up to two orders of magnitude higher than the geogenic sample rather pure effluent. As the water level of the Rhine River and background concentration. In particular, La concentrations observed its discharge were lowest in October 2009, the three samples (EF5, within an effluent plume in the vicinity of Worms are well-above EF6, and EF7) taken on 27.10.2009 should represent the purest those which have been found (e.g., Oral et al., 2010; Zhang et al., effluent, i.e. the effluent samples least diluted by normal Rhine River 2010) to cause ecotoxicological effects. The resulting anthropogenic water. positive La anomalies are further evidence that the increasing use of Tables 1 and 2 and Fig. 2 show the sampling locations, dates, and “exotic” elements, such as the REE, in high-technology products and related measured parameters for the samples presented here. The processes is accompanied by their increasing release into the concentrations present in the samples differ by several orders of environment. This highlights the urgent need for a better under- magnitude. Thus, for the sake of clarity, different units have been used standing of the distribution and biogeochemical behavior of the REE when reporting concentrations. in river, lake and ground water, in particular with regard to their ecotoxicology. 2.2. Methods

2. Sampling sites and methods Sampling was carried out using acid-cleaned 1000 mL polyethylene bottles. The analytical procedure has been described in detail elsewhere 2.1. Samples (Bau and Dulski, 1996; Kulaksız and Bau, 2007) and utilized a freshwater-adjusted version of the REE preconcentration method of Water samples from the Rhine River between Lake Constance in Shabani et al. (1992). All samples were passed through 0.2 μm filters and the south and the German–Dutch border in the north (Fig. 2A), and acidified to pH 1.8–2.0 with ultrapure HCl (required by the preconcen- additionally from specific sites (Fig. 2B) downstream from Worms tration method). Samples were then spiked with thulium (Tm) as an (between Rhine river-km 446 and 449; note that the official Rhine internal standard for monitoring preconcentration efficiency. Following river-kilometers system starts with river-km 0 shortly after Lake separation and preconcentration, REE concentrations were determined

Table 1 * * * Dissolved REE concentrations (ng/kg) in the Rhine River (GdSN/GdSN and LaSN/LaSN unitless). Tm interpolated using Er and Yb using Eq. (4) (see text).

Strasbourg Mannheim Pfrimm Worms- Worms- Worms- Worms- Mainz Lahnstein Bonn Leverkusen Neuss Xanten Wiembach 446.7 upstr. downstr. 448.8 Creek WWTP WWTP

Latitude 48.31209 49.43369 49.66065 49.66064 49.67417 49.67443 49.67845 49.97127 50.30040 50.74000 51.04346 51.18311 51.66470 51.07335 Longitude 7.71693 8.50634 8.36579 8.36607 8.35558 8.35534 8.35346 8.32867 7.60238 7.11326 6.94483 6.73364 6.48152 7.00322 Rhine river-km 446.7 446.7 448.4 448.45 448.8 Date 25.10.09 26.10.09 27.10.09 27.10.09 27.10.09 27.10.09 27.10.09 27.10.09 27.10.09 28.10.09 28.10.09 28.10.09 28.10.09 25.10.09 Time 16:40 16:38 9:20 9:30 11:20 11:25 11:45 12:52 16:54 11:36 14:19 15:45 20:00 12:00 T(oC) 13.2 14.3 11 15.4 15 17.2 15.1 13.6 13.3 12.8 13.8 13.4 12.5 11.9 pH 7.98 7.96 8.2 7.69 7.73 7.11 7.7 7.85 7.96 8.13 7.94 7.94 7.96 8.02 conductivity 385 419 1180 644 625 1500 584 540 594 589 647 701 725 601 (μS/cm) La 2.82 1.91 7.86 2.43 326 391 386 338 195 171 71.7 80.0 46.6 6.93 Ce 3.71 2.09 7.96 2.15 9.89 8.43 8.11 8.39 8.32 10.1 5.08 4.20 5.93 10.6 Pr 0.626 0.751 2.40 0.735 2.90 3.12 3.09 2.65 2.05 2.47 1.29 1.34 1.20 1.82 Nd 3.09 3.93 12.7 3.92 10.3 9.91 9.81 9.15 7.46 9.08 5.02 5.26 5.37 7.47 Sm 0.761 1.14 3.78 1.21 1.88 1.81 1.82 1.88 1.73 2.16 1.29 1.43 1.41 1.49 Eu 0.220 0.315 0.929 0.333 0.498 0.446 0.466 0.483 0.443 0.566 0.348 0.366 0.384 0.324 Gd 11.9 11.8 25.5 13.1 23.1 188 126 27.2 29.7 28.2 36.1 29.4 27.4 1.49 Tb 0.193 0.294 0.839 0.299 0.370 0.357 0.362 0.391 0.360 0.439 0.275 0.293 0.311 0.226 Dy 1.17 1.90 5.12 1.86 2.33 2.23 2.29 2.43 2.20 2.67 1.67 1.90 2.06 1.45 Ho 0.289 0.464 1.18 0.487 0.528 0.509 0.518 0.565 0.516 0.634 0.410 0.475 0.556 0.325 Er 0.896 1.51 3.76 1.59 1.73 1.67 1.71 1.90 1.78 1.97 1.52 1.70 2.00 1.01 Tm* 0.139 0.247 0.547 0.260 0.276 0.279 0.281 0.299 0.283 0.308 0.276 0.283 0.332 0.164 Yb 1.06 1.98 3.91 2.09 2.15 2.29 2.26 2.30 2.21 2.37 2.44 2.31 2.70 1.30 Lu 0.181 0.367 0.706 0.355 0.398 0.402 0.404 0.465 0.441 0.417 0.487 0.442 0.585 0.224 ΣREE 27.0 28.7 77.1 30.8 383 610 543 396 253 232 128 129 96.8 34.9 Gd* 0.96 1.7 5.8 1.9 1.8 1.7 1.7 2.0 2.1 2.6 1.7 2.0 1.9 1.5

Gdanthr 11 10 20 11 21 190 120 25 28 26 34 27 25 – * GdSN/GdSN 12 6.9 4.4 6.8 13 110 72 14 14 11 21 15 14 0.98

Gdanthr% 92857785929999939391959393 – La* 1.9 1.9 5.8 1.7 9.8 9.5 9.1 7.3 5.0 5.8 2.9 2.8 3.0 6.3

Laanthr 0.93 0.039 2.0 0.72 320 380 380 330 190 170 69 77 44 0.67 * LaSN/LaSN 1.5 1.0 1.3 1.4 33 41 42 46 39 29 25 29 16 1.1

Laanthr% 33 2.1 26 30 97 98 98 98 97 97 96 97 94 9.7

46 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979

Table 2 2.3. Quantification of REE anomalies Dissolved REE concentrations (μg/kg) in the effluent plume of a FCC cracking catalyst producing plant. Tm* data for EF1 interpolated using Er and Yb using Eq. (4) (see text). In semi-logarithmic plots the shale-normalized REE (REESN) EF1 EF2 EF3 EF4 EF5 EF6 EF7 patterns of natural substances are smooth, and neighboring REE can Date 28.01.09 5.5.09 5.5.09 5.5.09 27.10.09 27.10.09 27.10.09 be used for extrapolation or interpolation to calculate the natural Time 15:35 18:55 19:00 19:05 10:40 10:45 10:50 background values of those REE that are anomalously enriched or T (°C) – 17.2 ––29 –– depleted. In this paper the Gd anomaly was quantified using Nd and –– –– pH 7.04 6.9 6.95 Sm as discussed in detail elsewhere (Kulaksız and Bau, in press). Since conductivity 18,000 19,000 ––26,750 –– (μS/cm) La represents the beginning of the REE series, interpolation is La 16,600 27,000 13,500 14,900 48,800 49,300 48,300 impossible. Cerium cannot be used because it is redox-sensitive, and Ce 670 1410 584 648 2500 2550 2490 hence often behaves differently from the strictly trivalent REE. Pr 46.2 101 37.9 42.2 106 108 106 Therefore, the La background concentration is best determined by Nd 69.2 226 79.8 88.7 182 185 181 back-extrapolation using shale-normalized praseodymium (Pr) and Sm 5.77 18.5 6.35 7.03 12.7 13 12.6 Eu 0.721 1.57 0.591 0.664 2.01 2.01 1.99 neodymium (Nd): Gd 1.49 2.48 1.13 1.3 2.99 3.04 3.13  Tb 0.256 0.387 0.19 0.202 0.262 0.253 0.293 ⁎ ðÞ− ðÞ ðÞ log LaSN = 3log PrSN 2log NdSN 1 Dy 0.69 0.909 0.515 0.551 0.651 0.632 0.66 Ho 0.108 0.715 0.342 0.382 0.427 0.488 0.428 ⁎ Er 0.257 0.342 0.168 0.196 0.215 0.227 0.229 Where SN denotes shale-normalized concentration data and denotes Tm 0.656* 0.0105 0.00671 0.0106 0.0193 0.0175 0.018 calculated background values. The background concentration of La, Yb 0.11 0.146 0.0799 0.0894 0.163 0.166 0.176 La⁎, is calculated via: Lu 0.031 0.0204 0.0103 0.011 0.0421 0.0425 0.0441 ΣREE 17,400 28,800 14,200 15,700 51,600 52,200 51,100 ⁎ - ⁎ ½ ðÞ Cl (mg/L) ––––2800 2810 2830 La =LaSN ×LaPAAS 2 2- SO4 (mg/L) ––––17,400 17,400 17,400 where [La ] is the concentration of La in PAAShale of McLennan Samples taken at 49.66499°N, 8.36176°S, Rhine river-km 447.4 (Effluent-447.4 in PAAS Fig. 2A). (1989). ⁎ While the LaSN/LaSN ratio quantifies the La anomaly, the amount of excess La (later referred to as anthropogenic La, Laanthr) can be calculated as: by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). A small – ⁎: ð Þ subset of the samples, showing very high REE concentrations (EF2 to Laexcess =Lameasured La 3 EF7), was measured without REE separation and preconcentration (see Table 2), but was filtered and acidified following the routine sampling It is not possible yet to exactly quantify the usually small protocol described above. contribution of a natural positive La anomaly to a large anthropogenic For comparison, we included a sample from the lower reaches of one and it cannot be ruled out that waters which are rich in organic the Wiembach Creek, a small tributary to the Wupper River near the compounds and/or biota, for example, show analytically significant City of Leverkusen; the Wupper River itself is a tributary of the Rhine natural positive La anomalies. The sample labeled “Strasbourg” in River. This creek does not receive any input from WWTPs or industry, Fig. 3, for example, which was taken at a river section characterized and is considered a “pristine” control sample. We emphasize that this by extremely low flow velocity and large amounts of algae, shows a ⁎ sample possesses neither a Gd anomaly nor a La anomaly (Fig. 1A), LaSN/LaSN ratio of 1.5, suggesting that as much as 33% of its La might demonstrating that the anomalies we observe in other samples do not be anthropogenic (Table 1). Although this uncertainty in no way result from sampling or analytical artifacts. invalidates the results and general conclusions of our study, it

10 AB10 6

1 1 REE / PAAS x 10 REE / PAAS Mainz Lahnstein 0.1 0.1 Bonn Strasbourg Leverkusen Mannheim Neuss Worms-446.7 Xanten

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 3. REESN patterns of Rhine River water sampled at different locations. Significant positive anthropogenic Gd anomalies are visible in all samples (A and B) while severe La anomalies (and slight enrichment of Ce and Pr) are restricted to samples taken downstream of Worms (B).

47 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979 demonstrates that a more exact quantification of the anthropogenic La component requires a better understanding of the behavior of the ~ 50 mg/kg La 1000000 REE in organic- and biota-rich surface waters. In this study, we, Catalyst Plant Effluent (Jan. '09) therefore, arbitrarily (and conservatively) set a La /La⁎ ratio of N1.5 Catalyst Plant Effluent (May '09) SN SN Catalyst Plant Effluent (Oct. '09) as the lower limit before referring to a positive La anomaly as Pfrimm Creek anthropogenic. 100000 Rhine River: 3. Results and discussion Worms - 446.7 Worms - upstr. WWTP Σ Worms - downstr. WWTP Total dissolved REE concentrations ( REE) in our sample set of well- 10000 mixed Rhine River water vary from 27.0 ng/kg approximately 35 km Worms - 448.8 upstream from the City of Strasbourg, to over 396 ng/kg at the City of Mainz (Table 1); for the calculation of ΣREE, Tm⁎ concentrations are interpolated using Er and Yb as: 6 1000

∗ . . ð Þ Tm Tm 10 . 4

The REESN patterns (Fig. 3) show negative Ce anomalies of variable 100 size and display a general increase with increasing REE atomic number. REE / PAAS x 10 All samples show significant positive Gd anomalies, i.e., they carry anthropogenic Gd (Fig. 3) indicating widespread micropollution of 10 ~ 0.4 µg/kg La Rhine River water with Gd-based contrast agents. The sizes of the ⁎ Gd anomalies range from GdSN/GdSN =4.4 (Worms-446.7) to 21 (Leverkusen), revealing that between 85% and 95% of total dissolved Gd in the Rhine River is of anthropogenic origin. 1 In addition to these ubiquitous Gd anomalies, pronounced positive ⁎ La anomalies (LaSN/LaSN :16–46) are observed in all Rhine River samples taken downstream of river-km 447.4 (Fig. 3B), while all samples from upstream of this site (Fig. 3A) lack any significant La 0.1 ⁎ anomalies (LaSN/LaSN : 1.0–1.5). All samples that yield a positive La anomaly also show a small but consistent enrichment of Ce and Pr relative to the heavier REE (Fig. 3B). Note that this results in elevated Pr /Nd ratios which in turn produce La⁎ values calculated by Eq. 2 0.01 SN SN La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu that are slightly too high. The Laanthr values (Table 1) for the samples shown in Fig. 3B therefore represent minimum values. Fig. 4. Plot showing changes in REESN distribution between river-km 446 and 449, north fl Although small positive La anomalies may occur in seawater (e.g. de of Worms. Also shown is REESN data for the ef uent plume of an FCC catalyst plant at Baar et al., 1991) and, for example, result from partial scavenging of river-km 447.4 sampled at three different times during the year 2009. dissolved REE onto Fe oxyhydroxide surfaces (Bau, 1999; Bau and Koschinsky, 2009), the large size of the positive La anomaly in Rhine River stream within the discharge plume of the WWTP that serves Worms waters and the fact that large positive La anomalies have never been (population 82,000). These samples show REESN patterns that are very reported for waters from other rivers, strongly suggests that this positive similar to that seen upstream of the WWTP, with the exception of an ⁎ La anomaly is not natural but the result of anthropogenic La input. The increase in the anthropogenic Gd anomaly from GdSN/GdSN =13 ⁎ anthropogenic La component alone not only dominates the total (Worms-upstr.WWTP) to GdSN/GdSN =110 (Worms-downstr. dissolved La concentration but also the dissolved ΣREE content in the WWTP). This increase is due to additional input of anthropogenic Rhine River (e.g., 98% and 71%, respectively, in the sample Worms-448.8). Gd with the clear water discharge from the WWTP. As a result of the This is the reason for the significant difference between the dissolved continuous admixture of Rhine River water to the WWTP plume, ⁎ concentrations of ΣREE and La in samples from downstream and GdSN/GdSN then decreases downstream over a distance of about ⁎ upstream of Worms (Fig. 2). The sample taken at Mannheim (upstream 400 m from 110 to 72 (Worms-448.8). Interestingly, the LaSN/LaSN from Worms), for example, carries 1.91 ng/kg of dissolved La, while the ratio increases from 33 immediately upstream to 41 immediately sample from Mainz (downstream Worms) carries 338 ng/kg of dissolved downstream of the WWTP discharge site (Fig. 2B). This suggests that La. The downstream variation in the size of the positive La anomaly the clear water discharge of this WWTP carries a substantial amount indicates that anthropogenic La enters the Rhine River in the vicinity of of anthropogenic La. Unfortunately, this could not be verified, as we Worms, and that this signal is then diluted by La-uncontaminated waters did not get permission to sample the pure WWTP discharge. further downstream. This suggests that in marked contrast to anthropo- Samples taken at the effluent discharge pipe at km 447.4 show genic Gd, which is derived from a diffuse source (all WWTPs), the dissolved ΣREE concentrations of 52 mg/kg. For comparison, up- anthropogenic La is derived from a point source. stream of the discharge pipe and immediately before Pfrimm Creek In order to locate the actual site of anthropogenic input of La (and joins the Rhine River (sample Worms—446.7 in Fig. 4), the Rhine River to a lesser extent Ce and Pr) into the Rhine River, the area between carries only 30.8 ng/kg of dissolved ΣREE, i.e., roughly six orders of river-km 446 and 449(Fig. 2B) was sampled in more detail. As shown magnitude less. Since we could only sample the effluent immediately in Fig. 4, neither Pfrimm Creek, a small left-bank tributary to the Rhine after it is discharged into the Rhine River at km 447.4, our measured River, nor the Rhine River itself before km 447.4 shows significant maximum concentrations of dissolved ΣREE (52 mg/kg) and dissolved positive La anomalies. In marked contrast, a sample one kilometer La (49 mg/kg) still represent the respective minimum concentrations downstream from an effluent discharge pipe at km 447.4 but in the effluent. upstream of the discharge site of a WWTP shows a large La anomaly These anthropogenic concentrations are rather high compared to ⁎ (LaSN/La SN =33). Two additional samples were taken further down- typical dissolved concentrations of ΣREE and La found in pristine

48 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979 natural waters. While the dissolved La concentration in the effluent is occurred at La concentrations of 20 mg/kg (Chen et al., 2003). Zhang at least 49 mg/kg, the sample taken approximately one kilometer et al. (2010) recently reported on significant adverse effects on the downstream shows a La concentration of 320 ng/kg (Worms-upstr. growth and reproduction of worms (Caenorhabditiselegans)atLa WWTP), reflecting dilution of the effluent with relatively REE-poor concentrations above 1.39 mg/kg. The dissolved La concentrations of Rhine River water that does not show a positive La anomaly. up to 49 mg/kg we measured in the effluent plume of the point-source Nevertheless, the admixture of 1 L of effluent to 100,000 L of Rhine at river-km 447.4 are significantly higher, suggesting that this site River water is sufficient to produce the La concentration observed might provide a unique opportunity to study the ecotoxicity of downstream of river-km 447.4. Moreover, due to the extremely high anthropogenic La and its impact on biodiversity in a natural habitat. concentrations of La in the effluent plume relative to that of the Rhine River, the anthropogenic positive La anomaly still persists 400 km 4. Conclusions downstream to our northernmost sampling point at the City of – Xanten, close to the German Dutch border, where the size of This study of the REE distribution in the Rhine River in Germany fi ⁎ the positive La anomaly is still signi cant (LaSN/LaSN of 16). The documents the first observation of an anthropogenic La anomaly in concentration of dissolved La in this sample is 46.6 ng/kg, 94% of natural waters. The anthropogenic La is a microcontaminant that can which is of anthropogenic origin. Considering the discharge rate of be traced back to a FCC catalyst production plant, i.e., a point source, at 3 1080 m /s measured for the Rhine River at the Wesel gauge station on Rhine river-km 447.4, north of the City of Worms. Although derived the day of our sampling (W. Wiechmann, pers. com.) and conve- from a point source, the La microcontamination is still significant niently assuming that anthropogenic La discharge is constant over the (47 ng/kg of La, 94% of which is anthropogenic) some 400 km year, our data suggests an annual export of about 1.5 t of anthropo- downstream from this source, close to the German–Dutch border. genic Lavia the Rhine River into the North Sea. To put this number into The effluent from the catalyst plant contains exceptionally high perspective: annual world production of La oxide from 2012 to 2014 is dissolved total REE and La concentrations as high as at least 52 mg/kg predicted to range between 55,000 and 57,000 t (e.g., Kara et al., and 49 mg/kg, respectively. Since effects on organisms have been fl 2010). Hence, the 1.5 t of anthropogenic La ushed through the Rhine observed at La concentrations as low as 1.4 mg/kg (Zhang et al., 2010), River into the North Sea is substantial. the situation at the km 447.4 site should be monitored. However, it also fl The ef uent discharged into the Rhine River at km 447.4 originates offers the opportunity to study the ecotoxicological impact of La on a fl from a production plant for uid catalytic cracking (FCC) catalysts natural ecosystem exposed to elevated dissolved La concentrations. As fi fl used in petroleum re ning (note that in addition to the ef uent pipe the REE become increasingly more common in products and processes, to the Rhine River, the sewage system of this company is connected to it can be expected that in the near future increasing amounts of the public sewage system, which might explain the inferred presence anthropogenic REE will be released into the environment. Combined of anthropogenic La in the WWTP discharge). These FCC catalysts are with the observation that anthropogenic La and Gd are already present fi μ ne zeolite powders (average particle size of 60 to 100 m) in which as microcontaminants in some river and tap waters (e.g., Kulaksız and the REE are incorporated to increase stability at high temperatures Bau, in press), this trend strongly suggests that attention should focus (Yang, 2003; Kogel et al., 2006). Although such FCC catalysts are rich on REE distribution and behavior in different compartments of the fl in REE and show high LaSN/LuSN ratios similar to the ef uent at km natural environment, and that La and Gd should be closely monitored. 447.4 (Kulkarni et al., 2006), it is not likely that solid FCC catalysts are Moreover, geogenic REE background concentrations, particularly in discharged into the Rhine River, but rather water-soluble REE industrialized countries, need to be established before widespread compounds, such as REE chlorides, for example, that are used and contamination will have made this impossible. released during the production process. This is supported by the high concentrations of chloride (Cl−: 2800 mg/kg; Table 2) in the effluent. Acknowledgements Lanthanum is increasingly used in high-tech products and processes. In addition to FCC catalysts, La is not only applied with We thank G. Merschel, J. Mawick and D. Meissner (Jacobs REE-rich phosphate fertilizers in agriculture, but also used as additive University) for their help during sampling and lab work. We also in glass, in electron cathodes, scintillators, car batteries, dissolved thank W. Wiechmann for providing discharge data for the Rhine River. phosphate removal compounds, and batteries for hybrid cars, for This study was funded by Deutsche ForschungsGemeinschaft (DFG) example. Lanthanum carbonate is also used as a phosphate-binding through research grant BA 2289/2-1 to M.B. agent to treat hyperphosphataemia in renal failure and dialysis patients (Albaaj and Hutchison, 2005). Nevertheless, micropollution of the environment with La has so far only been observed in the References fi vicinity of petroleum re neries where it is related to solid La-rich dust Albaaj F, Hutchison AJ. Lanthanum carbonate (Fosrenol®): a novel agent for the particles (Kulkarni et al., 2006, 2007). To our knowledge, neither treatment of hyperphosphataemia in renal failure and dialysis patients. Int J Clin anthropogenic positive La anomalies in river water, nor such Pract 2005;59:1091–6. Bau M, Dulski P. Anthropogenic origin of positive gadolinium anomalies in river waters. extremely high dissolved La concentrations as those found within Earth Planet Sci Lett 1996;143:245–55. the effluent plume at Rhine River-km 447.4 have ever been reported Bau M. Scavenging of dissolved yttrium and rare earths by precipitating iron yet. Our data represent the first case of microcontamination of the oxyhydroxide: experimental evidence for Ce oxidation, Y-Ho fractionation, and lanthanide tetrad effect. Geochim Cosmochim Acta 1999;63:67–77. hydrosphere with dissolved La. Bau M, Knappe A, Dulski P. Anthropogenic gadolinium as a micropollutant in river While FCC catalyst production is identified as the cause of this waters in Pennsylvania and in Lake Erie, northeastern United States. Chem Erde anthropogenic La microcontamination, its effects on ecosystems in the 2006;66:143–52. river are largely unknown. Similar to other REE, La is a Ca inhibitor and Bau M, Koschinsky A. Oxidative scavenging of cerium on hydrous Fe oxide: evidence from the distribution of rare earth elements and yttrium between Fe oxides and Mn several adverse effects of La have been documented, such as decrease oxides in hydrogenetic ferromanganese crusts. Geochem J 2009;43:37–47. in the fertility of sea urchins (Oral et al., 2010). While Lacour et al. Ben-Dov IZ, Pappo O, Sklair-Levy M, Galitzer H, Ilan Y, Naveh-Many T, et al. Lanthanum (2005, 2007) reported a tenfold increase in La concentrations in body carbonate decreases PTH gene expression with no hepatotoxicity in uraemic rats. Nephrol Dial Transplant 2007;22:362–8. tissue of rats after oral administration of La carbonates, Ben-Dov et al. Byrne RH, Kim K-H. Rare earth element scavenging in seawater. Geochim Cosmochim (2007) and Damment et al. (2005) found no adverse effects in their Acta 1990;54:2645–56. studies. Slatopolsky et al. (2005) reported an accumulation of La in the Chen D, Liu Y, Chen A-j, Nie Y-X. Experimental study of subchronic toxicity of lanthanum nitrate on liver in rats. Nonlinearity Biol Toxicol Med 2003;1:469–80. livers of normal and uremic rats. While 6-month exposure to low- Damment SJP, Beevers C, Gatehouse DG. Evaluation of the potential genotoxicity of the dose La-nitrates was shown to promote growth in rats, liver damage phosphate binder lanthanum carbonate. Mutagenesis 2005;20:29–37.

49 S. Kulaksız, M. Bau / Environment International 37 (2011) 973–979

De Baar HJW, Schijf J, Byrne RH. Solution chemistry of the rare earth elements in Morteani G, Möller P, Fuganti A, Paces T. Input and fate of anthropogenic estrogens and seawater. Eur J Solid State Inorg Chem 1991;28:357–73. gadolinium in surface water and sewage plants in the hydrological basin of Prague Elbaz-Poulichet F, Seidel J-L, Othoniel C. Occurrence of an anthropogenic gadolinium (Czech Republic). Environ Geochem Health 2006;28:257–64. anomaly in river and coastal waters of Southern France. Water Res 2002;36:1102–5. Nozaki Y, Lerche D, Alibo DS, Tsutsumi M. Dissolved indium and rare earth elements in Gaillardet J, Viers J, Dupre B. Trace elements in river waters. In: Heinrich DH, Karl KT, three Japanese rivers and Tokyo Bay: evidence for anthropogenic Gd and In. editors. Treatise on geochemistry: surface and groundwater, weathering and soils. Geochim Cosmochim Acta 2000;64:3975–82. Oxford: Pergamon; 2003. p. 225–72. Oral R, Bustamante P, Warnau M, D'Ambra A, Guida M, Pagano G. Cytogenetic and Goldstein SJ, Jacobsen SB. Rare earth elements in river waters. Earth Planet Sci Lett developmental toxicity of cerium and lanthanum to sea urchin embryos. Chemo- 1988;89:35–47. sphere 2010;81:194–8. doi:10.1016/j.chemosphere.2010.06.057. Kara H, Chapman A, Crichton T, Willis P. Lanthanide resources and alternatives. UK: Pering C. (Gadopentetate dimeglumine) [PowerPoint slides]. Joint Meeting of the OakdeneHollins Research and Consulting; 2010. Available: http://www.oakdenehollins. Cardiovascular and Renal Drugs and Drug Safety and Risk Management Advisory co.uk/pdf/lanthanide_resources_and_alternatives_may_2010.pdf. Committee, Gaithersburg, MD.Bayer Healthcare Pharmaceuticals. USA; 2009. Available: Kim K-H, Byrne RH, Lee JH. Gadolinium behavior in seawater: a molecular basis for http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/ gadolinium anomalies. Mar Chem 1991:107–20. Drugs/CardiovascularandRenalDrugsAdvisoryCommittee/UCM196223.pdf. Knappe A, Möller P, Dulski P, Pekdeger A. Positive gadolinium anomaly in surface water Petelet-Giraud E, Klaver G, Negrel P. Natural versus anthropogenic sources in the surface- and ground water of the urban area Berlin, Germany. Chem Erde 2005;65:167–89. and groundwater dissolved load of the Dommel river (Meuse basin): constraints by Kogel JE, Trivedi NC, Barber JM, Krukowsk ST. Industrial minerals & rocks: commodities, boron and strontium isotopes and gadolinium anomaly. J Hydrol 2009;369:336–49. markets and uses. 7th. Littleton, CO, USA: Society of Mining, Metallurgy and Rabiet M, Brissaud F, Seidel J-L, Pistre S, Elbaz-Poulichet F. Deciphering the presence of Exploration; 2006. wastewater in a medium-sized Mediterranean catchment using a multitracer Kulaksız S, Bau M. Contrasting behaviour of anthropogenic gadolinium and natural rare approach. Appl Geochem 2005;20:1587–96. earth elements in estuaries and the gadolinium input into the North Sea. Earth Rabiet M, Brissaud F, Seidel JL, Pistre S, Elbaz-Poulichet F. Positive gadolinium Planet Sci Lett 2007;260:361–71. anomalies in wastewater treatment plant effluents and aquatic environment in Kulaksız S. and Bau M. Anthropogenic gadolinium as a microcontaminant in tap water the Herault watershed (South France). Chemosphere 2009;75:1057–64. used as drinking water in urban areas and megacities. Appl Geochem in press. Shabani MB, Akagi T, Masuda A. Preconcentration of trace rare-earth elements Kulkarni P, Chellam S, Fraser MP. Lanthanum and lanthanides in atmospheric fine in seawater by complexation with bis(2-ethylhexyl) hydrogen phosphate and particles and their apportionment to refinery and petrochemical operations in 2-ethylhexyl dihydrogen phosphate adsorbed on a C18 cartridge and determi- Houston, TX. Atmos Environ 2006;40:508–20. nation by inductively coupled plasma mass spectrometry. Anal Chem 1992;64: Kulkarni P, Chellam S, Fraser MP. Tracking petroleum refinery emission events using 737–43. lanthanum and lanthanides as elemental markers for PM2.5. Environ Sci Technol Slatopolsky E, Liapis H, Finch J. Progressive accumulation of lanthanum in the liver of 2007;41:6748–54. normal and uremic rats. Kidney Int 2005;68:2809–13. Lacour B, Lucas A, Auchere D, Ruellan N, de SerrePatey NM, Drueke TB. Chronic renal Tricca A, Stille P, Steinmann M, Kiefel B, Samuel J, Eikenberg J. Rare earth elements and failure is associated with increased tissue deposition of lanthanum after 28-day Sr and Nd isotopic compositions of dissolved and suspended loads from small river oral administration. Kidney Int 2005;67:1062–9. systems in the Vosges mountains (France), the river Rhine and groundwater. Chem Lacour B, Nikolov IG, Joki N, Drüeke TB. Lanthanum carbonate, body lanthanum Geol 1999;160:139–58. accumulation and potential liver toxicity. Nephrol Dial Transplant 2007;22:1788–9. United States Government Accountability Office. Rare earth materials in the defense Lawrence MG, Jupiter SD, Kamber BS. Aquatic geochemistry of the rare earth elements supply chain. GAO-10-617R; 2010. Available: http://www.gao.gov/new.items/ and yttrium in the Pioneer River catchment, Australia. Mar. Freshw. Res 2006;57: d10617r.pdf. 725–36. Verplanck PL, Taylor HE, Nordstrom DK, Barber LB. Aqueous stability of gadolinium in Lawrence MG. Detection of anthropogenic gadolinium in the Brisbane River plume in surface waters receiving sewage treatment plant effluent, Boulder Creek, Colorado. Moreton Bay, Queensland, Australia. Mar Pollut Bull 2010;60:1113–6. Environ Sci Technol 2005;39:6923–9. McLennan SM. Rare earth elements in sedimentary rocks; influence of provenance and Yang W-C. Handbook of fluidization and fluid particle systems. New York: Marcel sedimentary processes. Rev Mineral Geochem 1989;21:169–200. Dekker Inc.; 2003. Möller P, Dulski P, Bau M, Knappe A, Pekdeger A, Sommer-von Jarmersted C. Zhang H, He X, Bai W, Guo X, Zhang Z, Chai Z, et al. Ecotoxicological assessment of Anthropogenic gadolinium as a conservative tracer in hydrology. J Geochem Explor lanthanum with Caenorhabditis elegans in liquid medium. Metallomics 2010;2: 2000;69–70:409–14. 806–10. Möller P, Paces T, Dulski P, Morteani G. Anthropogenic Gd in surface water, drainage Zhu Y, Hoshino M, Yamada H, Itoh A, Haraguchi H. Gadolinium anomaly in the system, and the water supply of the city of Prague, Czech Republic. Environ Sci distributions of rare earth elements observed for coastal seawater and river waters Technol 2002;36:2387–94. around Nagoya City. Bull Chem Soc Jpn 2004;77:1835–42. Möller P, Morteani G, Dulski P. Anomalous gadolinium, cerium, and yttrium contents in Zhu Y, Hattori R, Rahmi D, Itoh SO, Fujimori E, Umemura T, et al. Fractional distributions the Adige and Isarco River waters and in the water of their tributaries (Provinces of trace metals in surface water of Lake Biwa as studied by ultrafiltration and ICP- Trento and Bolzano/Bozen, NE Italy). Acta Hydrochim Hydrobiol 2003;31:225–39. MS. Bull Chem Soc Jpn 2005;78:1970–6.

50 51 52 CHAPTER IV – NATURAL AND ANTHROPOGENIC RARE EARTH ELEMENT DISTRIBUTION OF THE RHINE RIVER AND ITS TRIBUTARIES

1. INTRODUCTION

Rare earth elements (REE) are comprised of lanthanum (La) and the elements cerium (Ce) to lutetium (Lu), a group of trace elements that behave coherently in natural systems. In natural waters, normalized REE concentrations plotted on a semi‐logarithmic scale reveal generally smooth patterns, except for Ce and europium (Eu), which may show deviation from the smooth pattern due to their redox chemistry. Cerium(III) can be oxidized to Ce(IV) and oxidatively scavenged, resulting in negative Ce anomalies in the dissolved phase, and, Eu(III) may be reduced to Eu(II) causing a positive Eu anomaly in high‐temperature low‐redox environments (e.g. hydrothermal waters). Other anomalies, usually only found in seawater, can be observed due to small differences in solution complexation stabilities of the individual REE, e.g. lanthanum, gadolinium and lutetium. Due to the similar ionic radii and identical charge of trivalent yttrium (Y) with that of the REE holmium (Ho), Y may be plotted together with REE in REY (REE and Y) plots. Despite being so‐called “geochemical twins”, Y can be decoupled from Ho and other neighboring REE to produce non‐chondritic Y/Ho ratios (Bau, 1996).

In natural waters, REE are present in three pools (Elderfield et al., 1990): particulate, colloidal, and truly dissolved. An operational filter size is defined (typically 0.2 µm or 0.45 µm) to separate the particulate REE pool from the dissolved pool. The filtrate (“dissolved” pool) contains the truly dissolved REE pool (free ions and solution complexes) and the colloidal REE pool. The truly dissolved natural REE pool is of great importance, because it is potentially bioavailable and can be thermodynamically modeled in terms of speciation. The colloidal REE pool is not easily removed from the truly dissolved REE pool due to the unavailability of reliable and cost‐efficient ultrafiltration or ultracentrifugation methods. As a result, data are commonly reported as “dissolved”, while in reality comprising of the truly dissolved REE pool and colloidal REE pool.

53 The colloidal load in river water constitutes a minor fraction of the total solid load, but can carry up to >90% of dissolved trace elements, such as REE, due to the strong association of the particle reactive REE with colloids (Sholkovitz, 1995). Thus, high‐colloid rivers have higher REE concentrations than rivers with low colloid content, while the removal of colloids accompanies the removal of REE from solution. Fractionation of REE also occurs between light REE (LREE) and heavy REE (HREE) due to the interplay between the HREE‐enriched truly dissolved load and the LREE‐enriched (or shale‐like) colloidal pool. Increasing HREE/LREE ratios with decreasing REE concentration also support this concept (Elderfield et al., 1990). Despite being particle reactive, REE can also form strong solution complexes with (in)organic ligands (e.g. Wood, 1990; Lee and Byrne, 1992). At low pH, free metal ions (Ln3+) and sulfate complexes control REE speciation, while in neutral to alkaline waters carbonate species dominate (Tang and Johannesson, 2003, and references therein).

Dissolved concentrations and distributions of REE in rivers are controlled largely by dissolved organic carbon (DOC) and pH. In general, decrease in pH causes increase in DOC, and consequently, increase in dissolved REE concentrations. The association of DOC and Fe with river colloids, and hence, of REE in rivers with DOC has been extensively documented (Sholkovitz, 1978; Sholkovitz et al., 1978; Elderfield et al., 1990; Sholkovitz 1993; 1995; Dupré et al., 1999; Ingri et al., 2000).

Research on the dissolved REE distributions in river waters (reviewed by, e.g., Elderfield et al., 1990; Sholkovitz, 1995; Gaillardet et al., 2003) shows a large variation of shale‐normalized patterns worldwide ranging from light REE (LREE) enriched to heavy REE (HREE) enriched. This observation is likely the result of the relative abundances of truly dissolved and colloidal loads in different settings. Elderfield et al. (1990) have suggested that a two‐component mixing between the colloidal and truly dissolved loads ultimately controls REE concentrations and distribution in river waters. Colloids carry a large fraction of REE and are characterized by LREE‐enriched REE patterns compared to the truly dissolved REE pool. The truly dissolved load on the other hand is HREE‐enriched and constitutes a minor fraction of the total REE. The higher the amount of colloids in a sample, the higher the REE content, and the less HREE‐ enriched the pattern (Elderfield et al., 1990; Sholkovitz 1995).

54 Currently, complete REE data covering large parts of the Rhine River are non‐existent, although data for limited geographies exists. Tricca et al. (1999) have studied dissolved and suspended REE concentrations from the Upper Rhine between Basel and Freiburg as well as in some tributaries of the Ill River in the Vosges Nantains. Dissolved neodymium (Nd) concentrations range between 2.6 and 7.4 ng/kg and shale normalized REE patterns are LREE depleted (0.12 < NdSN/YbSN < 0.23). A positive correlation between ytterbium (Yb) and DOC concentrations is reported with pH and source rock geology as the other factors suggested for controlling dissolved REE concentrations and distribution. Perret et al. (1994) have shown that colloids in the Rhine River provide a significant fraction of the available surface area for adsorption of pollutants, despite contributing less than 2% of the total particle volume and mass.

The current study focuses on natural and anthropogenic REE in the Rhine River, one of the largest rivers in Europe. With increasing anthropogenic pressure on rivers worldwide, the Rhine River is no exception. The riverbed has been changed numerous times to accommodate industrial needs in the past including the addition of dams and weirs to keep the river shippable even at low discharge. There has been continuous anthropogenic pressure on the Rhine River due to mining and agricultural activities in its catchment area, as well as intermittent (e.g. Sandoz chemical spill in 1986) sources of pollution that have affected the chemical and environmental balance of the river significantly (Milich and Varady, 1999).

Bau and Dulski (1996) first reported on anthropogenic REE with a study showing anthropogenic gadolinium (Gd) in the low discharge rivers of the Berlin area, as well as the Rhine River and its tributaries Wupper and Sieg. The source of the Gd anomaly was determined to be contrast agents used in magnetic resonance imaging, such as Gd‐DTPA and derivatives. These highly stable and water‐soluble compounds have since then entered surface waters through WWTP discharge (Bau and Dulski, 1996; Möller et al., 2002; Verplanck et al., 2005; 2010; Bau et al., 2006; Kulaksız and Bau, 2007; Lawrence et al., 2009; Petelet‐ Giraud et al., 2009; Rabiet et al., 2009), reported in groundwater (Möller et al., 2000; Knappe et al., 2005; Rabiet et al., 2009), wells used for drinking water production (Möller et al., 2000, 2002; Rabiet et al., 2005, 2006), tap water (Bau and Dulski, 1996; Möller et al., 2002; Kulaksız and Bau, 2011a), and coastal waters through estuaries (Kulaksız and Bau, 2007). The first

55 occurrence of positive anthropogenic Gd in the Rhine River was reported in the same study (Bau and Dulski, 1996), using a single sample taken at the city of Leverkusen.

In addition to Gd anomalies in the Rhine River, Kulaksız and Bau (2011b) have reported large anthropogenic La anomalies in the Rhine River downstream from Rhine‐km 447.3. The La contamination was traced to a fluid catalytic cracking (FCC) catalyst producing plant, which releases its waste or by‐product into the Rhine River. Dissolved La concentrations of up to 49 mg/kg have been measured In the vicinity of the discharge pipe, while La concentration approximately 50 km downstream at Rhine‐km 493 is 38 ng/kg, due to dilution. Nevertheless, La anomalies are significant even at Rhine‐km 823.3 close to the German – Dutch border.

The Rhine River receives waste water treatment plant (WWTP) effluents of 96% of the inhabitants of its catchment area (ICPR, 2005). While most pollutants have been largely removed from the effluents, micro‐pollutants that are harder to break down or remove from solution are not a part of routine removal processes in WWTP. Only few WWTP are efficiently able to remove anthropogenic Gd, such as the Advanced Water Treatment Plants (AWTPs) in Queensland, Australia (Lawrence et al., 2010). After going through the processes of microfiltration, reverse osmosis, and advanced oxidation (UV irradiation and peroxide), 99% of the anthropogenic Gd is removed, with reverse osmosis being responsible for the largest removal. For WWTP that lack reverse osmosis, anthropogenic Gd is not removed and discharge from WWTP effluent pipes continue to pollute surface waters. In addition to anthropogenic Gd, other pharmaceuticals are also likely to pass unharmed through WWTP treatment processes. In addition to anthropogenic Gd contamination via WWTP effluent, Weil and Knepper (2006) have reported contamination of the Rhine River with veterinary and human pharmaceuticals. Surface and ground water is contaminated via WWTP effluent due to infiltration or directly via animal excrement.

In order to further quantify the anthropogenic REE content in the Rhine River and its tributaries, information on the natural (background) REE content found in these waters is needed. An extensive study of natural and anthropogenic REE in the Rhine River currently does not exist. This study therefore investigates REE in the Rhine River and several of its tributaries (see Appendix 10 for a list) during three separate sampling campaigns starting

56 from approximately 30km downstream of Lake Constance, and ending at the German – Dutch border at Rhine‐km 857.3. A total of 156 samples were collected (79 of the Rhine River and 78 of its tributaries) roughly a third of which was taken in each sampling campaign. The motivation behind repeated sampling campaigns was to distinguish between two components of REE: background REE that are significant during high discharge periods, and anthropogenic REE that are significant during low‐discharge periods. This study supplements the growing body of literature on anthropogenic REE in the environment and provides necessary REE data for the Rhine River at a time when the amount of anthropogenic REE in the environment is rapidly increasing and REE occur as emerging microcontaminants in the hydrosphere.

57 2. SAMPLING AND METHODS

Surface waters of the main channel and tributaries of the Rhine River were collected during three sampling campaigns in May 2008, May 2009 and October 2009. Figure 2 in chapter I shows a map of the study area as well as selected sampling sites, while complete lists of sampling sites together with basic parameters can be found in the Appendices 3 – 8. Tributaries were sampled as close to the mouth as possible in order to acquire data that reflects the effect of the whole river on the Rhine River. In some cases this was not possible, due to urban constraints and because canals and alternative waterways from the tributary to the Rhine river complicated sampling. In these cases, samples were taken from the next closest upstream location.

For each sample a pre‐cleaned polyethylene bottle was used and rinsed several times using the surface water to be sampled. Electrical conductivity (EC) and pH of the rivers were measured in the field after sampling. The samples were kept in a cool and dark container until they were brought into the lab for processing in order to minimize biological activity.

2.1 SAMPLE PROCESSING AND ANALYSIS

The samples (1 L) were filtered through 0.2 µm filters before 10 mL aliquots were taken for analysis using ion chromatography (IC) for measuring major anions (filtered, non‐acidified). From the remaining sample volume, 100 mL aliquots were used for determining bicarbonate

‐ (HCO3 ) alkalinity by titration to pH 4.3 using 0.1 M HCl. The remaining sample volume was acidified to pH 1.8 – 2.0 before an aliquot of 20 mL was taken for measuring major cations using inductively coupled plasma optical emission spectrometry (ICP‐OES) analysis (filtered, acidified). After thulium (Tm) was added as an internal standard for checking recovery rates of the preconcentration method, another aliquot of 20 mL was taken for measuring rubidium (Rb), strontium (Sr), barium (Ba), Y and Tm using inductively coupled plasma mass spectrometry (ICP‐MS). The remaining sample volume was pre‐concentrated (approximately 100‐fold) for measuring REE (ICP‐MS), following the method of Bau and Dulski (1996) who adjusted the approach of Shabani et al. (1992) to optimize the pre‐concentration of REE in freshwater. Chapter 2 contains more details on the sample preparation and analytical procedures.

58 2.2 QUANTIFICATION OF ANOMALIES / EXTRAPOLATION OF MISSING DATA

On a semi‐logarithmic plot, normalized concentrations of elements La to Gd, and elements terbium (Tb) to Lu form two smoothly varying sub‐groups, with a step jump from Gd to Tb. This allows for the assumption that the elements La to Gd form one linear curve while the elements Tb to Lu form another. Thus, anomalies of elements within each of the two REE subgroups (La – Gd and Tb – Lu) can be calculated using elements from inside the same group.

The Rhine River carries large anthropogenic Gd and La (and to a lesser extent other LREE) anomalies as well as natural Ce anomalies. Additionally, Eu may potentially be anomalous, leaving Nd and samarium (Sm) as suitable elements for extrapolating background concentrations for La, Ce, Eu and Gd, all elements within the La – Gd subgroup. In order to calculate total dissolved REE (ΣREE), background Tm (Tm*) concentrations have also been extrapolated using the Tb‐Lu subgroup.

The following equations have been used for calculating background La, Ce, Eu and Gd concentrations.

∗ ∗ ∗ 5 3⁄2 eq. 1

∗ ∗ ∗ 2 eq. 2

∗ ∗ ∗ 3 ⁄2 eq. 3

∗ ∗ ∗ 2 eq. 4

59 * where signifies background concentrations and SN signifies normalization to PAAS (Post Archean Average Australian Shale; McLennan, 1989). For each element, the related background concentration and anthropogenic concentration can be calculated using:

∗ ∗ e.q. 5a

∗ e.q. 5b

where [RPAAS] is the concentration of R (any given REE) in PAAS.

* Using equations above, anomalies of La, Ce, Eu and Gd have been quantified (LaSN/La SN, * * * CeSN/Ce SN, EuSN/Eu SN and GdSN/Gd SN) as well as (extrapolated) natural background concentrations (La*, Ce*, Eu* and Gd*) and anthropogenic concentrations where appropriate

(Laanthr and Gdanthr). This process has been explained in greater detail in Appendix 11.

2.3 STUDY AREA

The main stem of the Rhine River is 1,233 km long (Milich and Varady, 1999) and drains over 2,000 m3/s of water to the North Sea. The long‐term average discharge (MQ) is 338 m³/s at the Konstanz gauging station (Germany), 1,260 m³/s at Karlsruhe – Maxau (Germany), and 2,270 m³/s at Rees, upstream of the German‐Dutch border (UN, 2007). Its catchment area over Austria, Belgium, France, Germany, Italy, Liechtenstein, Luxembourg, Switzerland, and the Netherlands is approximately 200,000 km2, with Germany constituting a little more than half of this total. Germany and the Netherlands host 37 million and 11.5 million residents of the 58 million total population that lives within the catchment area of the Rhine River, respectively. Between 20 (ICPR, 2005) and 30 million people in Europe receive drinking water from the Rhine River (www.iawr.org), while a substantial portion of the remaining population receives drinking water from tributaries of the Rhine River. The drinking water is supplied either by direct removal (e.g. Lake Constance) or via natural or artificial bank infiltration (e.g. several towns along the Ruhr area).

60 The main tributaries of the Rhine River are Aare, Main, Mosel and Neckar. The Aare River has a catchment area of 17,606 km2, has a yearly average discharge of 561 m3/s and hosts 3.4 million inhabitants in its catchment. The Main, Mosel and Neckar each have catchment areas of 27,251 km2, 28,133 km2, 13,950 km2, yearly average discharge of 193 m3/s, 328 m3/s, 145 m3/s and host populations of 6.6 million 4.3 million, 5.3 million within their catchments, respectively (Appendix 10).

2.3.1 Rhine River Sections

The Vorderrhein and Hinterrhein have their sources near Lake Toma and Paradies Glacier respectively, and join close to Reichenau to form the Alpine Rhine. While the Alpine Rhine (upstream of Lake Constance) constitutes a minor fraction (6%) of the total catchment area of the Rhine River (ICPR, 2005), it is responsible for a large portion of the total discharge. Especially in the summer, melt water from the Alps contributes significantly to the total discharge of the Rhine River (Hartmann et al., 2007). In the winter, precipitation runoff becomes dominant. The river then flows westward along the section known as the High Rhine (Hochrhein) between Lake Constance and Basel (Rhine‐km 165). The Aare River with its high discharge joins the Rhine River in this section. Downstream from the confluence of the Aare River and the Rhine River, the discharge rate in this section of the Rhine River is largely affected by several glacier and high mountain streams (Kempe and Krahe, 2005). The Upper Rhine (Rhine‐km 165 – 529) flows northward from Basel to Bingen with two major tributaries Neckar and Main joining the Rhine River in this section. It is situated along the major rift bearing the same name (Upper Rhine Graben) with the prevalent lithology being alluvial deposits (sandstone) from the Holocene and Pleistocene. Within the next section (Middle Rhine Rhine‐km 529 – 659) from Bingen to Bonn, the Moselle River joins the Rhine River. This section of the Rhine River shows lithologies of predominantly siliciclastic rocks (shales and claystones). Further downstream, the Lower Rhine (Rhine‐km 659 – 858) is covered by alluvial deposits from the Pleistocene and flows in wide meanders towards the German – Dutch border. The section downstream from the German – Dutch border to the North Sea (Delta Rhine) has a complicated hydrology with the varying interactions and watersheds between the Meuse, Ijssel and Rhine River systems depending on water levels. In order to keep sampling efforts limited to the Rhine River proper, alpine and delta sections of the Rhine River have not been sampled in this study.

61 2.3.2 Lake Constance

With its 536 m2 surface area, Lake Constance is the third largest lake in Central Europe. It is fed by several tributaries and forms the Rhine River proper at its outlet. Water from the lake is used for preparing drinking water for 4 million residents of Baden‐Württemberg. Theoretical residence time of the water in the lake is 4.3 years (KLIWA, 2012).

62 3. RESULTS

Appendices 3 – 9 show the results of the analyses for a range of elements and other parameters. Discharge data for each sampling location and date was obtained from the German Federal Institute of Hydrology (Bundesanstalt für Gewässerkunde – Wilfried Wiechmann, pers. comm.) using the available stations for monitoring hydrological data. In cases where no station close to sampling location was available, the next downstream station was chosen. Longitudinal discharge profiles for each sampling campaign are shown in Fig. 1. Sampling campaigns May 2008 and May 2009 are both characterized by high discharge while the October 2009 campaign shows low discharge, as is typical for this time of the year. Also shown are discharge values for the tributaries. The rivers Aare, Main, Moselle and Neckar show the highest discharge in that order for all three sampling campaigns.

2.5 May 2008 May 2009 October 2009

2.0 Emmerich Wesel Ruhrort Düsseldorf Köln Andernach Bonn Koblenz Kaub Mainz /s) 3 1.5 Worms Plittersdorf Maxau Philippsburg Speyer Helmlingen Grauelsbaum Basel-Rheinhalle Kehl-Kronenhof Hauenstein Kappel 1.0 Discharge (x1000 (x1000 m Discharge

0.5 Konstanz

0.0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 1. Discharge data for the day of sampling for each sampling campaign. The locations are labeled after the stations which record hydrological data (Wiechmann, pers. comm.).

63 3.1 MAJOR IONS

3.1.1 Rhine River

Major ion concentrations in the Rhine River are plotted against Rhine‐km in Figs. 2 to 11 and can be found in Appendix 4 and 5. For most elements there is a general increase while others show no clear trend. For the sake of readability, the charges of all ions have been omitted in this text.

200 Rhine River (May 09) 100 Rhine River (Oct 09) Tributaries (May 09)

Tributaries (Oct 09) Lippe 60 Erft

50 Mosel

40 Pfrimm Moder Main Nahe (mg/kg) + 30 Na Ruhr Wupper Neckar Sauer Sieg Wied Lahn Ahr 20 Töss Ill Wutach Acher/Rench Murg Kinzig 10 Wiese Aare Elz Thur Lauter 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 2. Sodium concentrations in the Rhine River and its tributaries against Rhine‐km.

Sodium (Na+) concentrations steadily increase with Rhine‐km (Fig. 2), from 2.3 – 3.7 mg/kg (May 2009 – Oct 2009) at Rhine‐km 34.4 to 22.3 – 34.7 mg/kg at Rhine‐km 823.3. There is a local Na+ peak of 12.3 – 29.26 mg/kg at Rhine‐km 446.7 with no significant effect on Na+ concentrations downstream from this location. The Rhine River before their confluence with the Lippe River (Rhine‐km 811.4) also shows a peak (25.4 – 38.3 mg/kg) in both sampling campaigns. Concentrations in the low discharge sampling campaign (October 2009) are consistently higher than those in the high discharge sampling campaign (May 2009).

64 Figure 3 shows chloride (Cl‐) concentrations in the Rhine River plotted against Rhine‐km. Concentrations increase steadily with two local peaks before the confluence of the Rhine River with the Pfrimm River (Rhine‐km 446.7) and the Lippe River (Rhine‐km 811.4). This is consistent with the Na+ data that shows elevated concentrations at the same locations. The sample from May 2009 also shows a peak at Rhine‐km 525.6 although the sample from October 2009 does not. Overall, Cl‐ concentrations vary from 7.16 – 12.2 mg/kg (May – October) at Rhine‐km 34.4 to 67.1 – 102 mg/kg at Rhine‐km 823.3.

550 Rhine River (May 09) Rhine River (Oct 09) 500 Tributaries (May 09) Tributaries (Oct 09) Lippe 260 240

220 Mosel 200 180 160

140 Pfrimm Erft (mg/kg) -

Cl 120

100 Moder Ill Main Töss

80 Sauer Neckar Nahe Wutach Wupper Ahr

60 Wied Kinzig Sieg Murg Acher/Rench

40 Wiese Thur Lauter Aare Elz 20 Lahn Ruhr 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 3. Chloride concentrations in the Rhine River and its tributaries against Rhine‐km.

Figure 4 shows a steady increase of calcium (Ca2+) concentrations in the Rhine River from 25.5 – 38.4 mg/kg at Rhine‐km 34.4 to 40.2 – 76.7 mg/kg at Rhine‐km 823.3. Data for May shows a steady increase while data from October 2009 shows strong increase in concentrations downstream of Rhine‐km 446.7. Similar to other major ions, concentrations in October are consistently higher than those in May.

65 Figure 5 shows a decrease in magnesium (Mg2+) concentrations at approximately Rhine‐km 100 followed by steady concentrations for the next 300 km. Downstream of Rhine‐km 400 there is steady increase in concentrations. Generally, Mg2+ concentrations increase from 4.36 – 7.67 (Rhine‐km 34.4) to 6.31 – 12.7 mg/kg (Rhine‐km 823.3) and concentrations in October are higher than those in May.

120 Rhine River (May 09) Rhine River (Oct 09) Tributaries (May 09) 100 Tributaries (Oct 09) Pfrimm Mosel

80 Neckar Main Töss Lippe Acher/Rench

60 Wutach Ill Moder Thur (mg/kg) Erft 2+ Nahe Ahr Ca 40 Wupper Sauer Wiese

20 Aare Lahn Ruhr Sieg Elz Lauter Kinzig Murg 0 Wied 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 4. Calcium concentrations in the Rhine River and its tributaries against Rhine‐km.

2‐ The Rhine River shows relatively small changes in sulfate (SO4 ) concentrations compared to 2‐ other major ions. Figure 6 shows SO4 concentrations generally vary from 36.6 – 60.8 mg/kg to 56.5 – 57.2 mg/kg for the Rhine River. Data for both May and October show similar trends, while samples upstream of Rhine‐km 300 show the largest difference. Concentration differences between the low discharge and high discharge sampling campaigns are highest in

2‐ this part of the river. Similar to other major ions, SO4 concentrations in the October samples are consistently higher than those in the May samples.

66

40 Rhine River (May 09) Rhine River (Oct 09) Tributaries (May 09) Tributaries (Oct 09)

30 Pfrimm Mosel Neckar 20 Erft Main Töss Ahr (mg/kg) Moder 2+ Nahe Mg Wupper Ill Sauer Wutach 10 Thur Lippe Elz Wiese Ruhr Kinzig Lahn Murg Acher/Rench Lauter Aare Sieg 0 Wied 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 5. Magnesium concentrations in the Rhine River and its tributaries against Rhine‐km.

Figure 7 shows potassium (K+) concentrations in the Rhine River steadily increasing with Rhine‐km from 0.72 – 1.85 mg/kg at Rhine‐km 34.4 to 1.92 – 6.48 mg/kg at Rhine‐km 823.3. K+ concentrations in October are higher than those in May and there is a pronounced increase at Rhine‐km 446.7 in the October samples.

‐ Figure 8 shows bicarbonate (HCO3 ) concentrations for the Rhine River, which are only available for the October sampling campaign. Concentrations increase sharply between Rhine‐km 304.6 and 354.3 from 61.5 to 107 mg/kg, and again between Rhine‐km 591.5 and 655.1 from 102 mg/kg to 183 mg/kg. There is also a local minimum at Rhine‐km 446.7 (73.1 mg/kg).

67

Rhine River (May 09) 180 Rhine River (Oct 09) Tributaries (May 09) Tributaries (Oct 09) 160 Wutach Mosel Neckar 140

120 Lippe Main 100 Moder Pfrimm Erft

80 (mg/kg) 2- 4 Ill

SO 60 Sauer

40 Murg Wiese Kinzig Aare 20 Ruhr Ahr Nahe Lahn Wupper Elz Sieg Thur Wied Lauter 0 Töss Acher /R 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 6. Sulfate concentrations in the Rhine River and its tributaries against Rhine‐km.

‐ Figure 9 shows nitrate (NO3 ) concentrations in May steadily increase with Rhine‐km from 4.02 mg/kg (Rhine‐km 34.4) to 6.92 mg/kg (Rhine‐km 823.3) while concentrations in October show a more complicated pattern. There is gradual increase from 4.85 mg/kg (Rhine‐km 34.4) to 10.72 mg/kg (Rhine‐km 304.6), followed by a sharp decrease to 5.55 mg/kg at Rhine‐km 354.3. From here, concentrations slowly increase further to 8.42 mg/kg (Rhine‐km 591.5) and

‐ fall rapidly to 2.77 mg/kg at Rhine‐km 655.1. Unlike the behavior of other major ions, NO3 concentrations in October are significantly higher than those in May only upstream of approximately Rhine‐km 300. Concentrations in October are slightly higher between approximately Rhine‐km 300 and 600, while concentrations in May are significantly higher

‐ downstream of Rhine‐km 600 than those in October. Data for NO3 is within the range of published data for the Rhine River at Lobith (Camusso et al., 2000: 3 – 5 mg/kg).

68 Figure 10 shows bromide (Br‐) concentrations in Rhine River decrease steadily with increasing Rhine‐km. Concentrations decrease from approximately 1.2 mg/kg at the High Rhine section, to approximately 0.2 – 0.3 mg/kg close to the German – Dutch border. Concentrations in samples from the most downstream location are comparable to published data (Wegman et al., 1981: 0.18 mg/kg at Lobith).

30 Rhine River (May 09) Rhine River (Oct 09) Tributaries (May 09) 25 Tributaries (Oct 09) Pfrimm

20

15 Nahe Erft (mg/kg) + Mosel K Main 10 Neckar Sauer Sieg Wupper Töss Moder Ahr Acher/Rench Kinzig Lippe Lauter Lahn 5 Wutach Ill Murg Thur Elz Wiese Ruhr Wied Aare

0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 7. Potassium concentrations in the Rhine River and its tributaries against Rhine‐km.

Figure 11 shows silicon (Si) concentrations steadily increasing against Rhine‐km, from 0.605 – 0.635 mg/kg at Rhine‐km 34.4 to 0.984 – 1.29 at Rhine‐km 823.3. There are several small deviations, most significant of which is at Rhine‐km 591.5 after the confluence of the tributary Lahn and the Rhine River, but before the confluence of the Mosel River and the Rhine River.

In October 2009, average total dissolved solids TDS = Ca2+ (51.0 mg/l) + Mg2+ (9.1mg/l) + Na+

+ ‐ ‐ 2‐ (19.0 mg/l) + K (3.9 mg/l) + Cl (51.2 mg/l) + HCO3 (92.9 mg/l) + SO4 (51.4 mg/l) + Si (1.02

69 220 Rhine River (Oct 09) Tributaries (Oct 09) 200 Erft

180 Pfrimm Thur Ahr 160

140 Main Neckar Töss

120 Lahn

100 (mg/kg) - Lippe 3 Nahe Sauer 80 Aare HCO Mosel Wupper

60 Moder Ill

40 Sieg Elz Wiese Acher/Rench Wied Ruhr

20 Wutach Kinzig Murg Lauter 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 8. Bicarbonate concentrations in the Rhine River and its tributaries against Rhine‐km.

mg/l) in the river is rather high (280 mg/l) compared with average river water (100 mg/l; Berner and Berner, 1996). The average dissolved Na+ / (Ca2+ + Na+) ratio of 0.25 is similar to that of most major rivers. This ratio, together with a TDS value of 280 mg/l, indicates dominance of the river water composition by carbonate – silicate weathering, rather than evaporation/precipitation, vegetation, or evaporate weathering (Gibbs, 1970). Samples from the Rhine River in October 2009 are above the 1 : 1 line on a plot of Ca2+ + Mg2+ (meq/l)

‐ 2‐ against HCO3 + SO4 (meq/l) and increases with Rhine‐km, suggesting that carbonate weathering is more significant for the cation control of the Rhine River as compared to silicate weathering (Fig. 12). This claim is further bolstered by the rather high Ca2+ + Mg2+ (meq/l) to Na+ + K+ (meq/l) ratios that are as high as 12.4 in our most upstream samples from the High Rhine section and sharply decrease to approximately 5 within 200 km. Equivalent Ca2+ + Mg2+ / Na+ + K+ ratios are above world average (2.2; Singh and Hasnain, 1999) for all samples upstream of approximately Rhine‐km 450, and fall below the world average ratio

‐ 2‐ ‐ 2+ upstream of Rhine‐km 493 (Fig. 13). Ternary diagrams of SiO2, HCO3 , SO4 + Cl and SiO2, Ca

70 + Mg2+, Na+ + K+ are routinely used for distinguishing different controls on water chemistry. These ternary diagrams have not been included here due to the extremely low amounts of Si found in our samples as compared to other major ions. Nevertheless, the average Si concentration in our dataset (1.0 mg/kg) falls within the range of Si concentrations reported elsewhere for the Rhine River (approximately 1 – 2 mg/kg in De Ruyter Van Steveninck, 1990; and 0.5 – 1.0 mg/kg in van Nieuwenhuyse, 2007).

70 Rhine River (May 09) Rhine River (Oct 09) Tributaries (May 09) 60 Tributaries (Oct 09)

50

40 Pfrimm Lippe

(mg/kg) 30 - 3 NO Neckar Töss Ahr

20 Main Ki

nzi Acher/Rench Ill Nahe Wutach

g Wupper Erft Moder Sieg

Elz Lahn Ruhr Wied Thur Wiese Sauer

10 Aare Lauter

Murg Mosel 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 9. Nitrate concentrations in the Rhine River and its tributaries against Rhine‐km.

Electrical conductivity (EC) of the Rhine River ranges from approximately 300 µS/cm in the most upstream location to approximately 600 – 700 µS/cm in our most downstream location (Fig. 14). Compared to published data (Buhl et al., 1991) EC values are somewhat lower. For instance, EC at Koblenz (Rhine‐km 591.5) is 501 µS/cm in the May 2008 sample, considerably lower than the value (714 µS/cm) reported by Buhl et al., (1991) but higher than the value of 390 as supplied by the German Federal Institute of Hydrology (15 May 2012, www.bfg.de). While there are small and rare deviations, the general trend is one of slow and steady

71 increase in EC with increasing river‐km in all sampling campaigns (May 2008, May 2009, and October 2009). The water samples presented here are slightly alkaline, with Rhine River pH ranging between approximately 7.8 and 8.4. In the two May campaigns pH gradually decreases with river‐km from approximately 8.4 to 8, while the October 2009 campaign shows no clear overall trend (Fig. 15). On average, pH is lower for the October 2009 sampling campaign (7.9) when compared to the May 2008 (8.2) and May 2009 (8.1) campaigns. Measured pH at Koblenz (8.12 in May 2008) is rather acidic when compared with available data: Buhl et al., (1991) and the German Federal Institute of Hydrology report pH values of 8.6 and 8.4 (15 May 2012, www.bfg.de) respectively.

3

Rhine River (May 09) Rhine River (Oct 09)

Töss Tributaries (May 09) Tributaries (Oct 09)

2 Thur Wutach (mg/kg) - Br 1 Elz Ill Moder Ahr Kinzig Sauer Wiese Pfrimm Neckar Lippe Lauter Mosel Main Lahn Nahe Erft Wupper Sieg Wied Ruhr Aare Acher/Rench Murg 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 10. Bromide concentrations in the Rhine River and its tributaries against Rhine‐km.

The majority of the samples are classified as soft (< 8.4 d°H) in terms of water hardness, with carbonate alkalinity and total alkalinity ranging from 2.1 d°H to 8.4 d°H (average: 4.3 d°H) and 6.7 d°H to 12.8 d°H (average: 9.2 d°H) respectively.

72 5 Rhine River (May 09) Rhine River (Oct 09) Tributaries (May 09) Tributaries (Oct 09) 4 Erft Lahn Pfrimm

3 Elz Sauer Lippe Lauter Main Wiese Moder 2 Acher/Rench Si (mg/kg) Neckar Nahe Wutach Ruhr Wupper Kinzig Sieg Wied Ahr Ill Murg Töss 1 Thur Mosel Aare 0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 11. Silicon concentrations in the Rhine River and its tributaries against Rhine‐km.

Figure 16 shows samples from the October 2009 campaign in a Piper plot. The samples

2‐ ‐ ‐ generally plot close to the SO4 + Cl line and show ion abundances in the order of HCO3 > 2+ 2+ ‐ + 2+ ‐ + 2+ 2+ SO4 > Ca > Cl > Na > Mg > NO3 > K . There is a weak evolution from more Mg + Ca + + 2‐ ‐ ‐ towards more Na + K while the change between the influence of SO4 + Cl and HCO3 is more pronounced. Concentrations of several elements (Rb, Ba, Si) and ions (Cl‐, Ca2+, K+, Na+) correlate with river‐km in both May 2009 and October 2009 sampling campaigns, while a clear decreasing trend is seen for Br‐. Magnesium data shows a weakly increasing trend, while U and Sr data show no clear trend with river‐km in both sampling campaigns although U levels are relatively higher within the first 100km of the Rhine River. For the sampling campaign of May 2008, Sr and U concentrations range between 296 – 403 (350) µg/kg and 0.29 – 1.09 (0.81) µg/kg, where numbers in parentheses indicate average concentrations. Additionally both sets of samples exhibit correlations between Na+ and Si (r2 = 0.67) and K+

2+ 2 2‐ 2+ and Mg (r = 0.68). While there is only a weak trend of SO4 increase with increasing Ca 2+ 2‐ 2 for the October 2009 campaign, Ca and SO4 also correlate well (r = 0.6) for the May 2009

73 campaign. The increase in Ca2+ and several other ions with increasing river‐km is in line with previously published data (Buhl et al., 1991). Nevertheless, there are also differences

‐ between the two sampling campaigns. For instance, NO3 data shows consistent increase with river‐km in the May samples, while no clear trend can be seen in the October samples.

2‐ Similarly, SO4 shows a weakly increasing trend in May but no clear trend in October. While there is increase in Cl‐ concentrations in the Rhine River between approximately river‐km 180 and 320 in both sampling campaigns, only data from the October 2009 campaign exhibits a pronounced local peak in this part of the Rhine River.

4.0 Rhine River (October 2009)

3.5

3.0 (meq/l)

2+ 2.5 + Mg 2+

Ca 2.0

1.5

1.0 1.0 1.5 2.0 2.5 3.0 - 2- HCO3 + SO4 (meq/l)

2+ 2+ ‐ 2‐ Fig. 12. Plot showing Ca + Mg concentrations (meq/l) against HCO3 + SO4 (meq/l). Color‐coding follows Rhine‐km, starting with red symbols at low Rhine‐km, via yellow, green and ending in blue at high Rhine‐km.

The Rhine River is classified as medium dilute (0.75 meq/l < Σ+ < 1.5 meq/l) to mineralized (3 meq/l < Σ+ < 6 meq/l) according to Meybeck’s (2004) classification scheme of water types (1.7 meq/l < Σ+ < 3.7 meq/l for the May 2009 samples and 2.8 meq/l < Σ+ < 6.0 meq/l for the October 2009 samples). According to this idealized scheme of water types based on a large database of world rivers, several factors are expected to affect the chemistry of the Rhine

74 River: Silicate weathering, carbonate weathering and rainfall. The low range of Σ+ (meq/l) for the Rhine river occurs mostly in samples from the High Rhine during the high discharge period (May 2009).

7 Rhine River (May 2009) Rhine River (October 2009) 6

5 ) (mg/kg) + + K + 4 ) / (Na ) / 2+ 3 World average = 2.2

+ Mg (Singh and Hasnain, 1999) 2+

(Ca 2

1

0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 13. Plot showing equivalent Ca2+ + Mg2+ to Na+ + K+ ratio against Rhine‐km. World average ratio of 2.2 from Singh and Hasnain, 1999.

3.1.2 The effect of the input of effluent at Rhine‐km 447.3

+ ‐ 2+ 2‐ There is a strong increase in concentrations of several major ions (Na , Cl , Ca , SO4 ) and EC in the Rhine River downstream of an effluent pipe discharging waste/processed water from an FCC catalyst producing plant at Rhine‐km 447.3. Samples within the plume of the effluent pipe itself show 3‐ to 5‐fold increase in Ca2+ and Mg2+ concentrations compared to samples

+ + ‐ 2‐ from the Rhine River immediately upstream, while concentrations for Si, K , Na , Cl and SO4 are 2 – 3 orders of magnitude higher. While the increase in concentrations for some ions is

2‐ still visible far downstream of the discharge pipe (most notably SO4 ), it should be noted that this part of the Rhine River is home to many industries which could provide an alternative

75 explanation for the marked increase in several of the major ions downstream of Rhine‐km 450.

23

22 Rhine River Tributaries 2.0 May 2008 May 2008 1.8 May 2009 May 2009 October 2009 October 2009 1.6 1.4 1.2 1.0 0.8 conductivity (mS/cm) 0.6 0.4 0.2 0.0 0 200 400 600 800 0 200 400 600 800 Rhine-km Rhine-km

Fig. 14. Conductivity data for the Rhine River and its tributaries.

9.2

9.0 Rhine River Tributaries May 2008 May 2008 8.8 May 2009 May 2009 October 2009 October 2009 8.6

8.4

8.2 pH 8.0

7.8

7.6

7.4

0 200 400 600 800 0 200 400 600 800 Rhine-km Rhine-km

Fig. 15. Measured pH for the Rhine River and its tributaries.

76 100 100 90 90 80 80 70

- 70 M Cl g 60 2 + + - 60 2 +

4 50 Ca

50 2 SO + 40 40 30 30 20 20 10 10 0 0

0 0 100 100

10 10 90 90

20 20 80 Na 80 - + 3 30 30 70 + 70

K + HCO SO + 40 40 2 60 60 g 4 2 - M 50 50 50 50

60 60 40 40

70 70 30 30

80 80 20 20

90 90 10 10

100 100 0 0 100 60708090 1020304050 0 0 102030405060708090100

Ca2+ Cl-

Fig. 16. Piper plot of the Rhine River samples from October 2009. The most upstream is shown with a green circle and the most downstream sample is shown with a red circle. The colors get darker with increasing Rhine‐km, and start with triangles, change to circles and finish with squares. The star signifies samples from an effluent pipe of a plant producing FCC catalysts at Rhine–km 447.3.

3.1.3 Tributaries

Figures 14 and 15 show coherent pH and EC values at the same location sampled at different times for the tributaries of the Rhine River. There are several tributaries that are consistently characterized by high EC values in all sampling campaigns (Fig. 14): Ahr, Erft, Ill, Lippe, Main, Moder, Mosel, Neckar, Sauer, Töss and Wutach, either due to high run‐off or industrial activity. Some of these tributaries show correspondingly high Na+ and Cl‐ concentrations (Figs. 2 and 3). Most notably, the Mosel River shows Na+ and Cl‐ concentrations of 58.8 mg/kg

77 and 247 mg/kg respectively in the October 2009 samples. The tributaries joining the Rhine River between approximately river‐km 180 and 320 drain the Alsace region of France, with its known potash mining industry. In contrast with the findings of Buhl et al. (1991), neither the Rhine River nor the tributaries in our dataset show peaks of Na+ or K+ concentrations in the tributaries of this region, although an increase in Cl‐ is observed in the October 2009 sample set. Thus, the impact of known anthropogenic pressures such as coal‐, salt‐ and potash‐ mining industries within the Alsace and Lorraine regions is only partially visible in our data. The Neckar, Main and Mosel rivers all drain catchment areas with abundant carbonate rocks,

2+ ‐ 2+ 2‐ which is reflected in the large proportion of Ca , HCO3 , Mg and SO4 (Figs. 4, 5, 6 and 8) ions in samples taken from these tributaries.

100 100 90 90 80 80 70 Wutach

- 70 l M C g 60 2 + Mosel + - Ill 60 2 +

4 50 C O Neckar a Moder 50 2 S LauterMainWieseKinzig + 40 TössElz PfrimmAcher/RenchSauerRuhr Aare WupperNahe 40 30 SiegWied Lippe 30 Murg 20 Lahn Thur Ahr 20 10 10 0 Erft 0

0 0 100 100

10 10 90 90

20 20 80 N 80 - a + 3 30 30 70 + O 70 C K + H 40 S + 40 Wutach 2 60 60 O g 4 2 - M 50 50 50 50

60 60 Neckar 40 40 Main Moder Ahr 70 70 Acher/RenchWiese 30 Aare ElzWupperRuhr Mosel 30 TössPfrimmLahn MurgPfrimmLauterKinzigIll WutachLauterMain WiedNaheErft 80 80 Sauer 20 ThurNeckarElz MoselWupperSieg Erft SiegWied 20 Aare IllWieseAcher/RenchSauerRuhrModer TössNahe KinzigMurg 90 90 AhrLahn 10 Lippe 10 Lippe Thur 100 100 0 0 100 2030405060708090 10 0 0 102030405060708090100

Ca2+ Cl-

Fig. 17. Piper plot for tributaries of the Rhine River (October 2009).

78

0.6 May 2008

0.5 /s) 3

0.4

0.3

0.2 Discharge (x1000 m

0.1

0.0

0.6 May 2009

0.5 Aare /s) 3

0.4

0.3

0.2 Main Discharge (x1000 m Mosel

0.1 Neckar Töss Thur Ruhr Lahn Sieg Lippe Wied Ahr Kinzig Nahe Erft Wiese 0.0 Murg Wupper

0.6 October 2009

0.5 /s) 3

0.4

0.3

0.2 Discharge m (x1000

0.1

0.0 0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 18. Discharge data for tributaries of the Rhine River.

79 3.1.3.1 Aare

With the highest discharge among the tributaries of the Rhine River, the Aare River is more voluminous (563 m3/s, yearly average) than the Rhine River itself (440 m3/s) before the confluence of these two rivers. The name Rhine is only used because the total length of its flow path is longer than that of the Aare River from source to the North Sea. The Aare River is dominated by melt water runoff in the Alps and is rather dilute (318 µS/cm in October 2009) compared with even the High Rhine (354 µS/cm) (Fig. 14). It is one of only a few tributaries that show major ion concentrations equal to or less than those found in the Rhine River. The

2+ ‐ only exceptions are Ca and HCO3 , which is likely due to the influence of carbonate weathering. Major ion concentrations are lower in the high‐discharge sample (Appendices 4 and 7 ‐ May 2009), probably as a result of dilution from melt water runoff during the high

2‐ ‐ discharge season. On a Piper diagram (Fig. 17), the Aare plots close to the SO4 + Cl line (i.e. very high Ca2+ + Mg2+ content).

3.1.3.2 Main

The Main River has the second highest discharge during all sampling campaigns and drains a large area with Triassic limestones, dolomites and impure carbonate sediments. As a result,

2+ 2+ 2‐ 2+ Ca , Mg , and SO4 concentrations are particularly high (e.g. in Oct 2009: Ca = 86.9 mg/kg, 2+ 2‐ 2+ 2+ Mg = 22.3 mg/kg, SO4 = 115 mg/kg) compared to the Rhine River (Ca = 96.0 mg/kg, Mg 2‐ = 9.4 mg/kg, SO4 = 49.4 mg/kg) before the confluence of the two rivers, as well as most of the other tributaries that drain different lithologies. Electrical conductivity of the Main River (870 µS/cm in October 2009) is also rather high compared to the Rhine River and other tributaries, with several other major ions also in the high range of our dataset (Na+, Cl‐, K+). The Main River plots close to the other tributaries on a Piper diagram (Fig. 17).

3.1.3.3 Mosel

The Mosel River has the third highest discharge during all three sampling campaigns, and drains lithologies of Upper Triassic impure carbonates and Middle Jurassic limestones (OneGeology Europe – http://onegeology‐europe.eu/). Carbonate weathering is expected to

2+ 2+ 2‐ largely dominate major ion chemistry of the Mosel River. Indeed, Ca , Mg and SO4 concentrations in the Mosel River are among the highest in this dataset (e.g. in Oct 2009: Ca2+

80 2+ 2‐ + ‐ = 102 mg/kg, Mg = 22.6 mg/kg and SO4 = 167 mg/kg; Figs. 4 – 6). Moreover, Na and Cl concentrations as well as EC values are also within the high end of the range of this dataset (in Oct 2009: Na+ = 58.8 mg/kg, Cl‐ = 247 mg/kg and EC = 1,218 µS/cm; Figs. 2, 3 and 14).

+ ‐ 2‐ 2+ 2+ ‐ Anthropogenic influence is greater on the ions Na , Cl and SO4 , while Ca , Mg and HCO3 concentrations are more stable in the face of anthropogenic pressure (Meybeck, 2004). The presence of soda industries that influence the Seille River, a tributary of the Mosel River, are the likely cause of its high EC and Na+ and Cl‐ concentrations. The Mosel River plots close to

2+ 2+ 2‐ ‐ the Mg + Ca line on a Piper diagram, with a high content of SO4 + Cl (Fig. 17).

3.1.3.4 Neckar

The Neckar River also has high discharge and drains significant areas with carbonates. All major ion concentrations in the Neckar River are higher (Figs. 2, 3, 4, 5, 6, 7, 8, 9 and 11) compared to the Rhine River upstream from their confluence. In particular, the biggest

2+ 2+ 2‐ differences are for Ca (94.5 mg/kg), Mg (19.2 mg/kg), and SO4 (141 mg/kg) compared to 2+ 2+ 2‐ the Rhine River (Ca = 45.7 mg/kg, Mg = 7.74 mg/kg and SO4 = 27.2 mg/kg, in October 2009; Figs. 4 – 6), a result of the carbonate weathering that dominates major ion chemistry in the catchment of the Neckar River. While exhibiting slightly higher Mg2+ + Ca2+ content, the Neckar River plots close to the other tributaries on a Piper diagram (Fig. 17).

3.1.3.5 Lippe

The Lippe River is a rather small river with a yearly average discharge of 45 m3/s. Nevertheless, it receives particular attention due its extremely high EC (0.18 mS/cm and 22.2 mS/cm in May 2008 and October 2009) and major ion composition (Figs. 2 – 11). All major ions in the Lippe River are significantly higher when compared to the Rhine River before their

‐ 2‐ confluence. For instance, in May 2009, Cl and SO4 concentrations were 592 mg/kg and 120 ‐ 2‐ mg/kg, compared to the concentrations of 126 mg/kg (Cl ) and 57 mg/kg (SO4 ) in the Rhine River before their confluence. Published major ion data for the Lippe River (Stögbauer et al., 2008) generally agrees with data presented here. The Lippe River plots far from the other

2+ 2+ 2‐ ‐ tributaries, close to the Mg + Ca line on a Piper diagram (i.e. very high SO4 + Cl content) and relatively high Na+ + K+ content (Fig. 17).

81 3.1.3.6 Wupper

The Wupper River is also a small tributary of the Rhine River, with a yearly average discharge rate of 15 m3/s. Major ion concentrations in the Wupper River are rather similar (for October

‐ 2‐ 2009, Cl = 46.5 mg/kg and SO4 = 42.7 mg/kg) to those in the Rhine River before their ‐ 2‐ 2+ ‐ confluence (Cl = 43.2 mg/kg and SO4 = 44.4 mg/kg), with the exception of Ca and HCO3 that are present in significantly lower concentrations in the Wupper River (Fig. 4 and Fig. 8). Like all tributaries downstream of Rhine‐km 520 (Middle Rhine and Lower Rhine), the

2+ ‐ Wupper River exhibits substantially lower Ca and HCO3 concentrations than found in the Rhine River, due to the disappearance of carbonates that are present in the catchments of tributaries in the Upper Rhine and High Rhine. Major ion concentrations for the Wupper River are within the range of published values (Stögbauer et al., 2008). The Wupper plots with the

2+ 2+ 2‐ ‐ other tributaries on a Piper diagram (Fig. 17) in the high Mg + Ca and high SO4 + Cl region.

3.2 RARE EARTH ELEMENTS

3.2.1 Rare Earth Elements in the Rhine River

All samples reported here exhibit normalized REE patterns that are HREE enriched. Since La exhibits a significant anthropogenic component in some samples and Lu could potentially be anomalous, quantification of the slope of REE patterns (LREE/HREE) was calculated using Nd and Yb. Normalized Nd to normalized Yb ratios (NdSN/YbSN) range between 0.20 – 0.43 (0.28), 0.13 – 0.36 (0.22) and 0.08 – 0.33 (0.21) for the May 2008, May 2009 and October 2009 sampling campaigns respectively, where numbers in parentheses are average values.

Total dissolved REE in the May 2008, May 2009 and October 2009 sampling campaigns average at 52.7, 59.9 and 111 ng/kg respectively. As will be discussed in detail under section 4.5 of this chapter, these total dissolved REE concentrations are often dominated by anthropogenic La and anthropogenic Gd. Dysprosium (Dy) is a purely geogenic REE in this dataset and hence is a good indicator of the geogenic levels of REE. Concentrations of Dy in the Rhine River average at 2.21, 1.89 and 1.69 ng/kg, and range between 1.02 – 4.44, 1.21 – 2.80 and 0.78 – 2.69, for the three sampling campaigns of May 2008, May 2009 and October 2009 respectively. In contrast, Gd, a largely anthropogenic REE, shows a much wider range of

82 concentrations (1.95 – 13.8, 1.81 – 15.3 and 9.12 – 51.3) and averages (6.13, 6.66 and 24.3) for the respective sampling campaigns of May 2008, May 2009 and October 2009. Lower total dissolved REE and higher anthropogenic REE (e.g. Gd) in the October 2009 sampling campaign is compatible with discharge data showing low discharge during this sampling campaign. In contrast, both May campaigns (2008 and 2009) show higher total dissolved REE and lower anthropogenic REE.

100

Rhine 34.4 km - May 08 Rhine 34.4 km - May 09 Rhine 34.4 km - Oct 09 Lake Constance -38 km - Aug 11 10 Lake Constance -38 km - Aug 11 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 19. Rare earth element patterns of the Rhine River at Rhine‐km 34.4. Two samples from Lake Constance are included (green symbols) for comparison.

A single sample and a duplicate from Lake Constance complement the dataset (August 2011). The two samples were taken independently of all three sampling campaigns and are not directly comparable in terms of natural REE with other samples. Nevertheless, the concentration of Dy in Lake Constance (0.93 ng/kg, Fig. 29) as well as the NdSN/YbSN ratio of 0.15 are close to those found in the next downstream sample during the low‐discharge season (Gailingen in October 2009: Dy 0.78 ng/kg, NdSN/YbSN = 0.17). This could possibly be due to the long residence time of water in Lake Constance (4.3 years), which has a stabilizing effect on seasonal variation in trace elements. In addition, Lake Constance shows a significant

83 * Gd anomaly (GdSN/Gd SN = 1.91 and 2.09, Fig. 19), an interesting finding, given that 4 million inhabitants in 320 cities of the state of Baden‐Württemberg use water from Lake Constance as drinking water (Zweckverband Bodensee‐Wasserversorgung, www.zvbwv.de), and the related efforts for preservation of the lake.

100

Rhine/Wupper 703 km - May 08 Rhine/Wupper 703 km - May 09 Rhine/Wupper 703 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 20. Rare earth element patterns of the Rhine River at Rhine‐km 703 (Leverkusen) before the confluence of the Rhine River and its tributary, Wupper River.

Rhine water samples at Gailingen (Rhine‐km 34.4) show a large variation both in natural and anthropogenic REE between sampling campaigns. Especially distinct is the October 2009 sample that shows rather low natural REE concentrations: Dy concentration in this sample is 0.78 ng/kg while both May samples have significantly higher Dy concentrations of 1.41 ng/kg (May 2008) and 1.21 ng/kg (May 2009). Additionally, the size of the Gd anomaly for both May

* * samples is small (GdSN/Gd SN = 1.96 in 2008 and GdSN/Gd SN = 1.17 in 2009; Fig. 19) while it is * extremely large in the October sample (GdSN/Gd SN = 38.6). The general REE pattern in all three samples is relatively constant with NdSN/YbSN ratios of 0.21, 0.15 and 0.17 for May 2008, May 2009 and October 2009 samples, respectively.

84 At Leverkusen (Rhine‐km 703), Dy concentrations of 3.26 ng/kg, 2.51 ng/kg and 1.67 ng/kg and Gd concentrations of 13.8 ng/kg, 15.3 ng/kg and 36.1 ng/kg were measured for the samples May 2008, May 2009 and October 2009 respectively. Normalized REE patterns are

LREE depleted, with NdSN/YbSN ratios of 0.32, 0.25 and 0.17 for the samples from May 2008, May 2009 and October 2009 respectively, and comparable with other samples in this river

* section. Both high discharge samples carry relatively small Gd anomalies (GdSN/Gd SN = 3.03 and 5.76 for May 2008 and May 2009 respectively; Fig. 20), while the low discharge sample

* * shows a GdSN/Gd SN ratio of 21.1. Similarly LaSN/La SN ratios of 8.5, 18.5 and 25.1 (May 2008, May 2009, October 2009) increase with decreasing discharge.

100

Rhine/Lippe 811.4 km - May 08 Rhine/Lippe 811.4 km - May 09 Rhine/Lippe 811.4 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 21. Rare earth element patterns of the Rhine River at Rhine‐km 811.4.

The most downstream location that is sampled in all three campaigns is at Rhine‐km 811.4, before the confluence of the Rhine River and its tributary the Lippe River. REE patterns of the two 2009 samples are highly coherent, while the May 2008 sample shows higher REE

85 especially for the LREE (Fig. 21). Dysprosium concentrations are 3.07 ng/kg, 2.00 ng/kg and 1.80 ng/kg for the May 2008, May 2009 and Oct 2009 samples respectively. Additionally, the

NdSN/YbSN ratio for the May 2008 sample is 0.24 while it is 0.13 and 0.14 for the May 2008 and May 2009 samples respectively. The size of the anthropogenic Gd at this location varies

* * between GdSN/Gd SN = 3.12 for the May 2008 sample and GdSN/Gd SN = 16.7 for the Oct 2009 * * sample, while the La anomaly varies between LaSN/La SN = 5.71 (May 2008) and LaSN/La SN = 13.2 (Oct 2009). The slope of the REE pattern from each sampling campaign is consistent with samples from the same sampling campaign further upstream: e.g. Rhine at Rhine‐km 779.4 shows NdSN/YbSN ratios of 0.30, 0.19 and 0.20 for the samples from May 2008, May 2009 and Oct 2009 respectively.

Data presented here is generally comparable to data from literature. Tricca et al. (1999) report Dy concentrations that range between 0.8 ng/kg and 2.4 ng/kg in the southern part of the Upper Rhine, with Gd concentrations that range between 1.3 ng/kg and 3.8 ng/kg. Data from the Rhine – Meuse estuary (Moermond et al., 2001) downstream from the German – Dutch border shows Dy concentrations ranging from 3.41 ng/kg to 46.2 ng/kg and Gd concentrations ranging from 3.30 ng/kg to 31.8 ng/kg.

3.2.2 Rare Earth Elements in the Tributaries

Samples from the tributaries of the Rhine River show average total dissolved REE in the May 2008, May 2009 and October 2009 sampling campaigns of 53.3, 66.6 and 68.7 ng/kg respectively. Dissolved Dy concentrations for the same sampling campaigns average at 2.60, 2.89 and 3.05 ng/kg, and range between 0.44 – 15.3, 0.48 – 11.1 and 0.69 – 13.8 ng/kg. The variation in the average and range of Dy concentrations in the tributaries is considerably smaller than that found in the Rhine River. In contrast to Dy, Gd shows much higher concentrations (20.6, 28.9 and 30.6 ng/kg) and variation (1.25 – 175, 1.21 – 314 and 5.08 – 77.4 ng/kg) in the tributaries. Similar to samples from the Rhine River, all tributaries are HREE enriched, with average NdN/YbN values of 0.22, 0.26 and 0.21, and span ratios of 0.08 – 0.63, 0.10 – 0.54 and 0.04 – 0.39. Some of the tributaries will be examined in more detail, on account of high discharge (Aare, Neckar, Main, Mosel) and unusual anthropogenic pressure (e.g. extremely high Gd anomaly in the Lippe and the Wupper).

86 3.2.2.1 Aare

100

Aare 103.2 km - May 08 Aare 103.2 km - May 09 Aare 103.2 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 22. Rare earth element patterns of the Aare River.

REE concentrations in the Aare River are significantly lower than those in the Rhine River and do not vary as much. For comparison, data from the May 2008, May 2009 and October 2009 campaigns show Dy concentrations of 0.72, 0.87 and 0.85 ng/kg in the Aare River, and 1.48, 1.46 and 1.12 ng/kg for the Rhine River upstream of their confluence. Figure 22 shows REE plots of the Aare River during different sampling campaigns. While discharge has significantly changed (Fig. 18) REE patterns have stayed relatively constant. The only striking difference is the markedly high anthropogenic Gd anomaly during the low discharge season (GdSN/GdSN = 10.2). This observation is shared by a majority of the samples from this dataset, both for the Rhine River and its tributaries. When compared to REE patterns of the Rhine River upstream of their confluence, the Aare River shows REE patterns that are flatter (NdSN/YbSN ratios of 0.23, 0.32 and 0.28, compared to 0.21, 0.18 and 0.19 for the Rhine River).

87 3.2.2.2 Main

Rare earth element plots of the Main River (Fig. 23) reveal a rather flat HREE section (Tb – Lu) with a steep LREE section (La – Gd), especially the two samples taken in the high discharge campaigns. This type of flat HREE sections in REE patterns are only found in two other tributaries in our dataset (Murg and Wutach, Figs. 3 and 13 of Appendix 2). In May 2009 the

Main River shows a TbSN/YbSN ratio of 0.83, while the Wutach and Murg rivers show TbSN/YbSN ratios of 0.92 and 0.97, respectively. Several samples taken from the High Rhine during May

2008 also show TbSN/YbSN ratios close to unity. For example, in May 2009 the Main River * carries higher concentrations of HREE (Yb = 3.02 ng/kg) and a large Gd anomaly (GdSN/Gd SN = * 6.30) compared to the Rhine River before their confluence (Yb = 1.80 ng/kg and GdSN/Gd SN = 3.75, Fig. 20 of Appendix 1 and Fig. 17 of Appendix 2). The Gd anomaly in the low discharge

* sample (GdSN/Gd SN = 20.3) is within the high range of this dataset.

100

Main 496.7 km - May 08 Main 496.7 km - May 09 Main 496.7 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 23. Rare earth element patterns of the Main River.

88 3.2.2.3 Mosel

100

Mosel 592.3 km - May 08 Mosel 592.3 km - May 09 Mosel 592.3 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 24. Rare earth element patterns of the Mosel River.

Dysprosium concentrations in the Mosel River are less than those found in the Rhine River before their confluence: 2.19 and 1.60 ng/kg in May 2008 and October 2009, when the Rhine River shows 1.60 and 2.10 ng/kg respectively. The Mosel River shows a steady LREE‐depleted coherent REE pattern (Fig. 24), with the two high discharge samples showing excellent agreement. The general pattern is also comparable to that of the Rhine River before their confluence (Fig. 23 in Appendix 1 and Fig. 20 in Appendix 2), although the Mosel River has steeper REE patterns (NdSN/YbSN ratios of 0.43 and 0.28 compared to NdSN/YbSN ratios of 0.20 and 0.11 for the Rhine River before their confluence, in May 2008 and October 2009. A

* moderate Gd anomaly is also present in the Mosel River (GdSN/Gd SN = 3.94 and 3.31) in May 2008 and May 2009, while the sample from October 2009 shows a more substantial anomaly of 15.0.

89 3.2.2.4 Neckar

Comparison of REE patterns for the Neckar River (Fig. 25) reveals consistent patterns for all three samples while the sample from May 2008 is comparatively low in REE concentrations. The reason for this is unknown, and is surprising based on the discharge data. The Neckar River shows higher concentrations of REE (e.g. Oct 2009, Dy = 2.68 ng/kg) than does the Rhine River (e.g. at Rhine‐km 414.3 in Oct 2009: Dy = 1.90 ng/kg). The Neckar River also

* carries a large anthropogenic Gd anomaly (e.g. Oct 2009: GdSN/Gd SN = 20.2), which does not seem to affect the anthropogenic Gd concentration in the Rhine River approximately 20 km

* downstream of their confluence (Sample “Rhine/Pfrimm” in October 2009: GdSN/Gd SN = 6.84; Fig. 33).

100

Neckar 428.2 km - May 08 Neckar 428.2 km - May 09 Neckar 428.2 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 25. Rare earth element patterns of the Neckar River.

3.2.2.5 Lippe

The Lippe River shows similar concentrations of REE (e.g. Oct 2009, Dy = 1.54 ng/kg) compared to the Rhine River (e.g. at Rhine‐km 811.4 in Oct 2009: Dy = 1.80 ng/kg), except in

90 May 2008 when the Rhine River shows elevated concentrations. REE patterns in the Lippe River (Fig. 26) are highly consistent between sampling campaigns and show rather steep slopes with NdSN/YbSN ratios of 0.08, 0.11 and 0.12 in the respective samples from May 2008, May 2009 and October 2009. The Lippe River also carries particularly high anthropogenic Gd

* anomalies (GdSN/Gd SN = 219 in May 2009 and 42.6 in October 2009). Although it has one of the largest Gd anomalies (May 2009) reported yet, the Lippe River does not change

* GdSN/Gd SN ratios in the Rhine River significantly due to its low discharge (Fig. 33). There is * also a small La anomaly in the October sample (LaSN/La SN = 4.26), although the presence of La anomalies in the tributaries of the Rhine River is unexpected.

100

Lippe 814.6 km - May 08 Lippe 814.6 km - May 09 Lippe 814.6 km - Oct 09

10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 26. Rare earth element patterns of the Lippe River.

3.2.2.6 Wupper

The Wupper River shows coherent REE patterns (Fig. 27) between different sampling campaigns with small changes in absolute concentrations in each sample. Variation in Dy concentrations is relatively small, with concentrations ranging between 1.09 ng/kg (October 2009) and 1.76 ng/kg (May 2009). Similarly, Gd concentrations are relatively constant and

91 range between 60.4 ng/kg (May 2008) and 84.1 ng/kg (May 2009). The Wupper River is a small tributary that receives relatively stable and high input of WWTP effluent (Iliev, 2010).

* Hence, GdSN/Gd SN ratios do not vary as much between sampling campaigns (37.7 in May 2008 and 64.5 in May 2009), especially when compared, for example, to the Töss River (2.26 – 66.0) or the Mosel River (3.31 – 15.0). The second sample from May 2009 is a duplicate (light blue in Fig. 27) and shows good agreement with the sample from May 2009 (dark blue).

100

Wupper 703.4 km - May 08 Wupper 703.4 km - May 09 Wupper 703.4 km - May 09 Wupper 703.4 km - Oct 09 10 6

1 REE / PAAS x 10 x / PAAS REE

0.1

0.01 La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 27. Rare earth element patterns of the Wupper River.

92 4. DISCUSSION

4.1 NATURAL RARE EARTH ELEMENTS

100

Murg Main Rhine River: Murg 10 May 2008 Wiese May 2009 MurgPfrimm Acher/Rench October 2009 Main ModerKinzigKinzig PfrimmWiese Elz ElzIll Tributaries: ModerSauer Acher/Rench LauterWutachNeckarNahe May 2008 RuhrMain NaheIll RuhrSauerWiese SiegWiedNeckarLippe Mosel

Dy (ng/kg) Ill May 2009 Wutach KinzigAhrWupperModerWutach Mosel Ahr Elz NaheThur October 2009 Acher/Rench Thur WiedMoselNeckar Lahn LauterRuhr LippeLippeErft LahnWupperSieg Aare 1 Ahr Wupper World Rivers (Gaillardet, 2003) Sieg Birs Lahn Erft Thur Rhine River (Tricca et al., 1999) Aare Aare WiedErft Töss Töss

Töss

456789pH

Fig. 28. Plot showing Dy vs. pH for the Rhine River and its tributaries. Data for world rivers (Gaillardet et al., 2003) and the Rhine River (Tricca et al., 1999) included for comparison.

REE concentrations in the Rhine River reported here are comparable to published data. Figure 28 shows Dy concentrations for the Rhine River and its tributaries against pH, together with data for world rivers compiled by Gaillardet et al., (2003). Several rivers included (e.g. Rio Negro, Zaire River, Nyong River) are rich in DOC and have pH values below 6, resulting in high REE content in these waters. The Rhine River and its tributaries have rather high pH and low REE as compared to world rivers. Samples presented here group close to Rhine River data (green stars in Fig. 28), together with other rivers such as the rivers Mississippi, Mackenzie and St. Lawrence. The inverse relationship of REE concentrations with pH is well known (Debert et al., 2002; and references therein). Ingri et al. (2000) have shown a 7‐fold increase in La concentrations as a result of increase in discharge and DOC, and decrease in pH during spring flooding. For Dy data presented here for the Rhine River and its tributaries, there is also a tendency of decreasing Dy with increasing pH. However, there are some tributaries

93 with exceptionally high Dy concentrations (e.g. Murg, Main, Wiese) despite having pH values not significantly lower than other rivers.

In general, REE concentrations in the Rhine River (e.g. Dy concentrations; Fig. 29) increase with Rhine‐km in all sampling campaigns. This result is not surprising given that pH in the

Rhine River decreases with increasing Rhine‐km (Fig. 15). Figure 30 shows NdSN/YbSN ratios for the Rhine River and its tributaries against pH together with data for world rivers (Gaillardet et al., 2003). Data from this study shows high pH low NdSN/YbSN ratio and groups with the Rhine

River (green star symbol; Fig. 30). There is a weak trend of decreasing NdSN/YbSN ratios with increasing pH within especially the May 2008 and May 2009 samples, while no clear trend can be seen for the October 2009 samples.

20 May 2008: May 2009: October 2009 18 Rhine River Rhine River Rhine River Tributaries Tributaries Tributaries August 2011: 16 Lake Constance

14

12

10 Dy (ng/kg)

8

6

4

2

0 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

Fig. 29. Dysprosium concentrations for the Rhine River and its tributaries against Rhine‐km.

4.2 CE ANOMALY

The oxidation of Ce(III) to the insoluble Ce(IV) leaves the solution phase depleted in Ce(III) leading to negative Ce anomalies. Sholkovitz (1995) suggested that this process is promoted

94 by high pH and the abundance of surface sites (colloids), and that rivers with a high pH, low colloid and high carbonate content show low concentrations of dissolved REE, strongly HREE‐ enriched normalized patterns and large Ce anomalies (Sholkovitz, 1995). Removal of colloids from solution leaves the dissolved pool with a more negative Ce anomaly (Sholkovitz, 1992). Hence, the colloidal pool carries positive Ce anomalies relative to the truly dissolved REE pool, and the abundance of colloids should influence the size of the Ce anomaly.

1

Rhine River: May 2008 May 2009 SN October 2009 /Yb

SN Tributaries: May 2008 Nd 0.1 May 2009 October 2009

World Rivers (Gaillardet et al., 2003) Rhine River (Tricca et al., 1999)

456789 pH

Fig. 30. Plot showing NdSN/YbSN ratios for the Rhine River and its tributaries. Data for world rivers (Gaillardet et al., 2003) and Rhine River (Tricca et al., 1999) included for comparison.

Aside from its high pH and low REE concentrations, the Rhine River also shows HREE enrichment and significant Ce anomalies. All but a few samples from the Rhine River show

* negative Ce anomalies ranging between CeSN/Ce SN = 0.39 (May 2008, Rhine‐km 393.9) to * CeSN/Ce SN = 0.85 (May 2008, Rhine‐km 313.2). The size of the Ce anomaly does not vary systematically with Rhine‐km. Positive Ce anomalies appear in samples with exceptionally

* steep LREE slopes (e.g. for May 2009, Rhine‐km 34.4: CeSN/Ce SN = 1.34). Positive Ce anomalies are typically found in alkaline lakes (Möller and Bau, 1993; Johannesson et al.,

95 1994), and anoxic seawater (Bau et al., 1997; Schijf et al., 1991; 1995). While not common, positive Ce anomalies have also been reported for river waters (Elderfield et al., 1990).

1 SN * Rhine River:

/Ce May 2008

SN May 2009

Ce October 2009 Tributaries: May 2008 May 2009 October 2009

World Rivers (Gaillardet et al., 2003) Rhine River (Tricca et al., 1999)

0.1 456789 pH

* Fig. 31. Plot showing CeSN/Ce SN ratios against pH for the Rhine River and its tributaries. Data for world rivers (Gaillardet et al., 2003) and Rhine River (Tricca et al., 1999) included for comparison. For * consistency, Ce SN for literature data has been recalculated according to eq. 2.

A large percentage of the tributaries also carry negative Ce anomalies ranging from

* * CeSN/Ce SN = 0.34 (Ahr River, May 2008) to CeSN/Ce SN = 0.86 (Lauter River, May 2009), while a * few also carry positive Ce anomalies and CeSN/Ce SN ratios close to unity. For example, the * Acher/Rench sample exhibits CeSN/Ce SN ratios between 0.92 (Oct 2009) and 1.33 (May 2008), * while the rivers Moder, Lauter, Lahn, Ahr and Erft show CeSN/Ce SN ratios close to unity (0.97 * < CeSN/Ce SN < 1.04). While there is no conclusive explanation for the presence of positive Ce anomalies in some samples in this dataset, it is worth mentioning that most of the tributaries

* that exhibit CeSN/Ce SN ratios close to unity were taken during the low‐discharge campaign in October 2009. One possible explanation may be the relatively low pH values (Fig. 15) and scarcity of colloids in this sampling campaign. This explanation is similar to that suggested by Sholkovitz (1993): the presence of particles and colloids promote Ce(III) removal, and

96 conversely, in the absence of particles and colloids, Ce does not undergo fractionation. While

* Elderfield et al. (1990) have shown an inverse relationship between CeSN/Ce SN ratios and pH, this observation does not hold for data presented here (Fig. 31).

1 SN * /Ce SN

Ce Rhine River: May 2008 May 2009 October 2009 Tributaries: May 2008 May 2009 October 2009

0.1 World Rivers (Gaillardet, 2003) Rhine River (Tricca et al., 1999)

1 10 100 1000 Nd (ng/kg)

* Fig. 32. Plot showing CeSN/Ce SN ratios against Nd concentrations for the Rhine River and its tributaries. Data for world rivers (Gaillardet et al., 2003) and the Rhine River (Tricca et al., 1999) included for * comparison. For consistency, Ce SN for literature data has been recalculated according to eq. 2.

Previously, all but few authors reporting Ce anomalies have either used La and praseodymium (Pr), or La and Nd, to interpolate background Ce concentrations (Ce*). As argued in Appendix 11, this approach is not fundamentally sound and in the case of the Rhine River, impractical due to the presence of large anthropogenic La anomalies. A direct

* comparison of reported CeSN/Ce SN ratios from other rivers with data presented here is * therefore not possible. Nevertheless, CeSN/Ce SN ratios recalculated from literature data according to eq. 2 range between 0.20 and 1.47 and have an average close to unity (0.94).

* The Rhine River shows CeSN/Ce SN ratios of 0.58, 0.69 and 0.67 for the sampling campaigns of May 2008, May 2009 and October 2009. Figure 32 shows Ce anomalies against Nd concentrations (as a representative LREE) in the Rhine River and its tributaries together with

97 * world rivers. An interesting trend is the decrease in CeSN/Ce SN ratios with increasing Nd concentrations for the Rhine River samples. There is no explanation offered regarding this point.

Shiller (2002) also observed a seasonal influence of microbial activity on Ce anomalies, where samples taken in the summer showed on average more negative Ce anomalies, due to increased microbial Mn oxidation providing fresh absorbing surfaces for LREE (Shiller, 2002).

* Data from the Rhine River shows no marked difference in CeSN/Ce SN ratios between sampling campaigns (Fig. 32).

4.3 ANTHROPOGENIC GD

May 2008:

May 2009: Lippe October 2009: Rhine RIver Rhine River Rhine River Tributaries Tributaries Tributaries August 2011: 100 Lake Constance Wupper SN * /Gd SN Lahn Gd Ruhr Nahe Sieg

10 Neckar Main Wied Pfrimm Töss

1

0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

* Fig. 33. Plot showing GdSN/Gd SN ratios (eq. 4) for the Rhine River and its tributaries.

All water samples taken from the Rhine River as well as from its tributaries during the three sampling campaigns carry positive anthropogenic Gd anomalies. Anthropogenic Gd

* anomalies (GdSN/Gd SN) increase steadily with increasing Rhine‐km during both May

98 campaigns (2008 and 2009), especially downstream of approximately Rhine‐km 400 (Fig. 33). While showing a similar behavior of increase with Rhine‐km, samples from the October 2009

* campaign show additional features: GdSN/Gd SN ratios are 2 – 3 times higher, and there are * several prominent peaks when compared with the May campaigns. Furthermore, GdSN/Gd SN ratios are exceptionally high in the High Rhine and parts of the Upper Rhine (up to

* approximately Rhine‐km 300). The size of the Gd anomaly (GdSN/Gd SN) ranges between 1.41 – 3.41 (2.22), 1.16 – 8.01 (3.35), and 5.33 – 61.8 (17.11) for the sampling campaigns of May 2008, May 2009 and October 2009.

40 May 2008: May 2009: October 2009: Rhine River Rhine River Rhine River Tributaries Tributaries Tributaries August 2011: Lake Constance 30

20

(ng/kg) nat Gd 10

0 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

Fig. 34. Plot showing natural Gd (Gdnat) concentrations for the Rhine River and its tributaries as calculated by eqs. 4 and 5.

The Rhine River already carries a Gd anomaly at Lake Constance (Fig. 19) however the size of the anomaly is further increased at approximately Rhine‐km 65. With the exception of the Töss River (Rhine‐km 73.4) there is no significant input of anthropogenic Gd from tributaries in this section of the Rhine River. The additional anthropogenic Gd therefore likely results

99 from the relatively low discharge at this section of the Rhine River and the specific time of the year coupled with input from WWTP effluents between Rhine‐km 34 and 65. Of the several WWTP found in this section of the Rhine River, the WWTP in Neuhausen (am Rheinfallen) serves seven municipalities with a total population of 50,000 inhabitants (hwww.abfall‐ sh.ch/roeti/) and discharges into the Rhine River at Rhine‐km 49. The anthropogenic Gd anomaly in the Rhine River also peaks at approximately Rhine‐km 628 with no significant input from two tributaries joining the Rhine River after the last sample of the Rhine River (Rhine‐km 591.5). The WWTP at Koblenz is located at Rhine‐km 595 and treats domestic wastewater for more than 100,000 inhabitants as well as commercial and industrial wastewater.

300 May 2008: May 2009: October 2009: Rhine River Rhine River Rhine River 200 Tributaries Tributaries Tributaries August 2011: Lake Constance

60

(ng/kg) anthr Gd

20

0 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

Fig. 35. Plot showing anthropogenic Gd (Gdanthr) concentrations for the Rhine River and its tributaries as calculated by eqs. 4 and 5.

Together with high discharge tributaries Neckar, Main, Sieg and Ruhr, some smaller tributaries such as Lippe, Wied, Lahn and Nahe also show significantly high anthropogenic Gd anomalies, in some cases much higher than those observed in the Rhine River.

100 4.3.1 Variation of anthropogenic Gd with weekday

Rhine River: May 2008 May 2009 October 2009 100 Tributaries: May 2008 SN

* May 2009 October 2009 /Gd SN Gd

10

1

Sat Sun Mon Tue Wed Thu Fri Sun Mon Tue Wed Thu Sun Mon Tue Wed

* Fig. 36. Plot showing GdSN/Gd SN ratios in the Rhine River against weekday on which each sample was collected.

The size of the Gd anomaly depends, among other things, on the number of applications of Gd‐based contrast agents, which in turn depend on operating times of MRI facilities. There is a daily fluctuation of Gd concentrations in hospital effluent (Kümmerer and Helmers, 2000) and a weekly fluctuation depending on the day of the week (Bremen Seehausen WWTP, unpublished data). The utilization of MRI facilities peaks during weekdays and daytime, while it is at a minimum over the weekend and nighttime. Depending on the time it takes sewage to reach the WWTP and the residence time within the WWTP, the minima and maxima in Gd

* input undergo delays before they can influence the surface water. As a result, GdSN/Gd SN ratios are influenced by the day of the week sampling is carried out in surface waters. The

* effect is likely to be bigger for small rivers that receive WWTP effluent with high GdSN/Gd SN ratios. For studies of anthropogenic Gd in individual rivers, this point should be taken into account. Unfortunately, as completing each sampling campaign within one day was not possible, weekly variation of Gd anomalies cannot be ruled out in this sample set. At average

101 discharge, the time it takes a water parcel from Lake Constance to reach Lobith (Rhine‐km 865) is 12 days (De Ruyter Van Steveninck, 1990), implying that the variation in anthropogenic Gd anomalies with weekday should be negligible compared to the variation with Rhine‐km.

Figure 36 shows the size of the Gd anomaly against weekday for each sampling campaign. All sampling campaigns started on a Sunday, but ended on a Friday, Thursday and Wednesday for the May 2008, May 2009 and October 2009 campaigns respectively. There is a tendency for smaller Gd anomalies in the Rhine River and its tributaries on Mondays and Tuesdays compared to Thursday or Friday, especially during the low discharge sampling campaign. A similar plot of Dy against weekday shows no consistent trend, suggesting that the minimum in Gd anomalies at the beginning of the week is due to the weekly variation in anthropogenic Gd input.

100 Rhine River: May 2008 May 2009 October 2009

10 SN * /Gd SN Gd

1

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

* Fig. 37 ‐ Plot showing GdSN/Gd SN ratios against discharge for the Rhine River.

102 * Similarly, GdSN/Gd SN ratios increase as the week progresses, although this could be an effect due to the presence of more tributaries in the Lower Rhine area that receive higher inputs of anthropogenic Gd through WWTP effluents due to the population density of their catchment areas. The samples taken on the last days of each campaign are from the area known as the Ruhr‐area in Germany, a conglomerate megacity that meets the prerequisites for producing high Gd anomalies in surface waters: high population density and number of MRI applications, and small rivers that receive WWTP effluent contaminated with anthropogenic Gd (e.g. the rivers Wupper, Ruhr and Lippe). Variation of Gd with weekday is an expected observation and is supported in Fig. 36, but should be received with caution, and should be verified with daily sampling of the same location in future studies.

May 2008 34.4 May 2009 811.4 October 2009 703 779.4 103 493 584.5 735.4

10 525.6

304.6 SN 354.3 *

/Gd 655.1 SN 294.2 Gd

1

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

* Fig. 38. Plot showing GdSN/Gd SN ratios against discharge for individual sampling locations on the Rhine River. Labels indicate the Rhine‐km of each sampling location.

In addition to the input of anthropogenic Gd, discharge also plays a significant role in changes

* in GdSN/Gd SN ratios. Samples of the Rhine River taken during the low discharge sampling * campaign show unusually high GdSN/Gd SN ratios that show a weakly increasing trend with discharge. During this period of low discharge, the High Rhine section has considerably lower

103 runoff than during the high discharge period, and Gd input is not diluted as much. In contrast, samples from the high discharge campaign show significantly smaller Gd anomalies due to the apparent dilution effect from increased discharge. However, and more interestingly, there are increasing trends with discharge for the samples from May 2008 and May 2009 (Fig. 37). The increase in Gd anomalies with discharge is not expected; due to a decrease in the relative influence of anthropogenic Gd input from WWTP effluents with the addition of rainwater and runoff. There is a confounding factor that affects anthropogenic Gd anomalies and seems to correlate with discharge and Rhine‐km. At the High Rhine, elevation is high (high run‐off) and population density is low, decreasing the effect of anthropogenic Gd input with high‐discharge tributaries (e.g. Aare). As Rhine‐km (and Rhine discharge) increases, the Rhine River flows through more densely populated regions and tributaries with highly contaminated anthropogenic Gd. Direct WWTP effluent discharge into the Rhine River increases correspondingly in areas with higher population density. Hence, the effect of increase in Gd anomalies with Rhine‐km (and discharge) within each sampling campaign is due to the elevated input of Gd, which overshadows the effect of dilution i.e. as Rhine‐km increases, the increase in anthropogenic Gd input is higher compared to the increase in discharge.

60

55 May 2008: May 2009: October 2009: Rhine River Rhine River Rhine River 50 Tributaries Tributaries Tributaries 45 August 2011: Lake Constance 40

35 SN * 30 /La SN 25 La

20

15

10

5

0 0 200 400 600 800 10000 200 400 600 800 10000 200 400 600 800 1000 Rhine-km

* Fig. 39. Plot showing LaSN/La SN ratios (eq. 1) for the Rhine River and its tributaries.

104 The effect of discharge on Gd anomalies can be more appropriately visualized using plots of

* GdSN/Gd SN ratios against discharge at each sampling location (Fig. 38). Variation in population density and input of Gd at a given sampling location are negligible compared with the variation between sampling locations. Thus, with increasing discharge, the size of the Gd anomaly decreases. Discharge plays a two‐fold role in the size of Gd anomalies. As discharge increases, background (natural) REE concentrations also increase, resulting in higher Gd*

* (background Gd) concentrations and lower GdSN/Gd SN ratios. Moreover, the addition of rainwater and runoff causes a decrease in the influence of anthropogenic Gd input from

* WWTP effluents due to dilution, further increasing GdSN/Gd SN ratios.

40 May 2008: May 2009: October 2009: Rhine River Rhine River Rhine River Tributaries Tributaries Tributaries August 2011: Lake Constance 30

20

(ng/kg) nat La 10

0 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

Fig. 40. Plot showing natural La (Lanat) concentrations for the Rhine River and its tributaries as calculated by eq. 1.

4.4 ANTHROPOGENIC LA

* Figure 39 shows LaSN/La SN ratios in the Rhine River against Rhine‐km. Variation of anthropogenic La with Rhine‐km is very different from that of anthropogenic Gd. Except for small peaks at Rhine‐km 154 and 318.5 in the May 2008 sampling campaign, there is no

105 significant anthropogenic La upstream of Rhine‐km 447.3. Anthropogenic La anomalies increase sharply by two orders of magnitude downstream of waste water effluent input from a factory producing Fluid Catalytic Cracking (FCC) catalysts at Rhine‐km 447.3. The point‐ source contamination from this effluent is gradually diluted by the addition of water from tributaries that in general lack any anthropogenic La. Nevertheless, anthropogenic La anomalies are significantly high even in the most downstream samples almost 400 km downstream from the site of contamination. This behavior is observed in all sampling campaigns and is very similar in both high‐discharge campaigns (May 2008, May 2009). During low‐discharge sampling (October 2009), the anthropogenic La peak is much sharper and more pronounced due to the reduced effect of dilution from additional waters with no anthropogenic La. Moreover, the peak for Laanthr occurs further downstream in each of the two high discharge campaigns (Rhine‐km 525.6), while Laanthr peak for both the low discharge campaign (October 2009) shows a peak at Rhine‐km 493.

350 May 2008: May 2009: Rhine River Rhine River 300 Tributaries Tributaries August 2011: Lake Constance October 2009: Rhine River 250 Tributaries

200

(ng/kg) anthr La 100

50

0

0 200 400 600 800 Rhine-km 0 200 400 600 800

Fig. 41. Plot showing anthropogenic La (Laanthr) concentrations for the Rhine River and its tributaries as calculated by eq. 1.

The effect of discharge on La anomalies can be clearly observed in Fig. 42, which shows

* LaSN/La SN ratios against discharge at each sampling location for all three sampling campaigns.

106 * At each location, LaSN/La SN ratio decreases with increasing discharge, consistent with the point source nature of the La contamination. The seasonal fluctuation in discharge is more significant than the variation in the input of anthropogenic LREE (especially La) into the Rhine River.

4.5 ANTHROPOGENIC VS. NATURAL RARE EARTH ELEMENTS

525.6 493 735.4

584.5 655.1

703 779.4

10 811.4 SN * /La SN 34.4 La 304.6

103

354.3 May 2008 1 294.2 May 2009 October 2009

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

* Fig. 42. Plot showing LaSN/La SN ratios against discharge for individual sampling locations on the Rhine River. Labels indicate the Rhine‐km of each sampling location.

Figures 29, 39 and 34 show measured Dy, interpolated natural La (Lanat), and also interpolated natural (Gdnat) against Rhine‐km. There is no clear difference between sampling campaigns in May (2008 and 2009) and the sampling campaign in October (2009) in terms of

Dy, Lanat or Gdnat concentrations. The general increase in Dy concentrations with Rhine‐km is well mirrored by natural Gd concentrations. In addition to this behavior, natural La

107 concentrations strongly increase downstream of Rhine‐km 447.3. This is due to the interpolation of background La anomalies (eq. 1) using Nd and Sm, since Nd is also affected to some extent by the contamination at Rhine‐km 447.3. In marked contrast to these three

“natural” concentrations, the variation in anthropogenic La (Laanthr) and anthropogenic Gd

(Gdanthr) between the May and October sampling campaigns is clearly visible (Figs. 35 and 41). During low discharge (October) the anthropogenic elements are even more elevated than they are during high discharge (May). Furthermore, the samples from the October campaign are more prone to local changes while the May samples are comparatively more stable over the course of the Rhine River.

500 Rhine River: Rhine River: Rhine River: May 2008 May 2009 October 2009: Total REE Total REE Total REE Natural REE Natural REE Natural REE 400 Lake Constance: August 2011 Total REE Natural REE REE (ng/kg) REE

200

100

0 0 200 400 600 800 0 200 400 600 800 0 200 400 600 800 Rhine-km

Fig. 43. Plot showing total REE (filled symbols) and natural REE (open symbols) against Rhine‐km.

Due to the presence of significant amounts of anthropogenic REE, variation in the background (natural) REE is often masked. This point can be clearly seen in Fig. 43 that shows total dissolved REE against Rhine‐km together with natural REE against Rhine‐km. In each of the sampling campaigns, there are significant deviations from the two lines, especially with increasing Rhine‐km. The variation in the abundance of natural REE is distinctively different from that of total REE. Total natural REE concentrations rarely go above 50 ng/kg while total

108 REE can be as high as 500 ng/kg. The behavior and controlling factors are also entirely different. While concentrations of natural REE vary with seasonal discharge, anthropogenic REE are controlled by local phenomena and do not show significant seasonal variation as compared to natural REE.

4.5.1 Anthropogenic La vs. Gd

16 Rhine River: May 2008 14 May 2009 October 2009

12

10

8 Nd (ng/kg) 6

4

2

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

Fig. 44. Plot showing Nd concentrations against discharge for the Rhine River.

The behavior of anthropogenic La and anthropogenic Gd are completely different from one another. Anthropogenic Gd originates from Gd‐based contrast agents used in MRI scans. After near‐total excretion from the body after administration, these contrast agents pass through WWTPs largely unaffected, eventually landing in surface waters via the effluent pipes of WWTP. In this regard, each WWTP that discharges into a stream or river constitutes a source of contamination for anthropogenic Gd. In marked contrast, the anthropogenic La contamination originates from a point‐source. The production of REE‐doped zeolites that are used as FCC catalysts results in large amounts of REE, especially La but also LREE, in the effluent of this particular plant as a waste or by‐product. Within the plume and the immediate vicinity of the discharge pipe, La concentrations as high as 49 mg/kg have been

109 measured (Kulaksız and Bau, 2011b). As a result of the point‐source nature of the

* anthropogenic La contamination, LaSN/La SN ratios progressively decrease downstream of this location to 46.3 at the next sampling location 46 km downstream from the site of contamination, and to 14.5 further 330 km downstream. Figures 39, 41 and 42 illustrate the effect of dilution on the anthropogenic La anomaly between approximately Rhine‐km 450 to our most downstream sample at approximately Rhine‐km 820.

5 Rhine River: May 2008 May 2009 October 2009 4

3 Yb (ng/kg) Yb 2

1

0 200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

Fig. 45. Plot showing Yb concentrations against discharge for the Rhine River.

Discharge during May 2008 is the highest while discharge during May 2009 was slightly lower (Fig. 1). Discharge during the low discharge sampling campaign (October 2009) was approximately half that of the high discharge sampling campaigns. With a few exceptions, sampling sites within each campaign show increase in discharge with Rhine‐km. There is a trend of increasing REE concentrations with increasing discharge (Fig. 44), likely as a result of higher input of colloidal REE. If this is the case, LREE and HREE should behave differently with varying discharge i.e. LREE that are enriched in the colloidal pool should be more severely affected by changes in discharge while HREE should be less affected.

110 1 SN /Yb SN Nd

Rhine River: 0.1 May 2008 May 2009 October 2009 Lake Constance: August 2011

0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 46. Plot showing NdSN/YbSN ratios against Rhine‐km for the Rhine River.

There is excellent correlation among the LREE (e.g. r2 = 0.96 for Pr vs. Nd) and the HREE (e.g. r2 = 0.97 for Ho vs. erbium (Er)). On the other hand, the correlation between LREE and HREE is weak (r2 = 0.51 for Nd vs. Yb), suggesting that LREE and HREE are controlled by different sources and processes. At individual sampling locations, HREE show less seasonal variability compared to LREE: For instance, Rhine at Leverkusen (Rhine‐km 703) shows a two‐fold increase in Nd concentrations between the low discharge season and the high discharge season, while Yb concentrations only increased by approximately 10%. Additional data from Leverkusen (Table 1, Chapter V) shows a linear relationship (r2 = 0.77) between discharge and

NdSN/YbSN ratios, highlighting the influence of colloids not only on the REE concentrations but also on REE distribution at this location. Figure 48 shows the variation of NdSN/YbSN ratios at Leverkusen (Rhine‐km 703) and other sampling locations along the Rhine River. Although there is no direct relationship with discharge, there is a tendency for higher NdSN/YbSN ratios during the May 2008 campaign (high discharge). Data for Leverkusen (Table 1, Chapter V) shows increased scatter at low discharge (approximately 1,000 m3/s) with a variability of 0.3

111 in NdSN/YbSN ratios. Thus, the low NdSN/YbSN ratios (Fig. 48) observed for the May 2009 campaign are within the range of variation.

1 Wiese Wutach Elz Kinzig Murg Thur Birs Main Pfrimm Neckar SN Acher/Rench Lahn /Yb Wied Lauter SN Nahe Sauer Nd Erft Ahr Aare Ruhr Ill Töss Wupper Lippe Sieg

0.1 Tributaries: Moder

May 2008 Mosel May 2009 October 2009

0 100 200 300 400 500 600 700 800 900 Rhine-km

Fig. 47. Plot showing NdSN/YbSN ratios against Rhine‐km for the tributaries of the Rhine River. Please note that filled symbols are used to denote tributaries for legibility.

Figure 46 shows maximum NdSN/YbSN ratios for each sampling campaign at approximately Rhine‐km 150 and Rhine‐km 600 while minimum ratios are observed between Rhine‐km 300

– 350, forming an “M‐type” double convex down pattern. Tributaries show NdSN/TbSN ratios (Fig. 47) that vary similarly, with maximum ratios at approximately Rhine‐km 150 and minimum ratios at approximately 300 – 350. Downstream of approximately Rhine‐km 300,

NdSN/YbSN ratios are consistently higher than those in the Rhine River, suggesting that input from tributaries has an influence on the dissolved REE distribution of the Rhine River. The effect of catchment geology could explain the variation in the Middle and Lower Rhine. The Middle Rhine has a prevalent lithology of shales and claystones. As a result, the Rhine River and its tributaries within the Middle Rhine show elevated NdSN/YbSN ratios. In the Lower

112 Rhine, NdSN/YbSN ratios decrease back to levels observed in the Upper Rhine, due to the change in lithologies from shales in the Middle Rhine to sandstones in the Lower Rhine.

May 2008 May 2009 October 2009 0.4

525.6 493

0.3 655.1

SN 584.5 304.6 /Yb SN

Nd 779.4 294.2

0.2 103 735.4 34.4

703

811.4 0.1 354.3

200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

Fig. 48. Plot showing NdSN/YbSN ratios against discharge for individual sampling locations on the Rhine River. Labels indicate the Rhine‐km of each sampling location.

A comparison of Nd (Fig. 44) and Yb (Fig. 45) concentrations in the Rhine River against discharge reveals that both elements show significant variability. Within each subset of samples, Nd concentrations form a concave‐down curve as concentrations reach maximum values in the mid‐discharge section of the Rhine River. On the other hand, Yb concentrations show a more linear relationship with increasing discharge. Hence, the variation of NdSN/YbSN ratios with discharge is controlled by Nd concentrations and the behavior is similar to that of Nd: concave‐down with maximum ratios occurring in the mid‐discharge section (Fig. 49).

There is no consistent variation in average NdSN/YbSN (0.28, 0.22 and 0.21) and TbSN/YbSN (0.72, 0.61 and 0.63) ratios between the high discharge and low discharge sampling

113 campaigns. Both ratios show a concave‐down curve when plotted against discharge.

Moreover, there is no consistent change in NdSN/YbSN and TbSN/YbSN ratios with pH, similar to results by Shiller (2002), where the author reported no major change in REE concentrations and distributions with pH, suspended load or organic complexation for the Mississippi River.

Rhine River: May 2008 0.4 May 2009 October 2009

0.3 SN /Yb SN Nd 0.2

0.1

200 400 600 800 1000 1200 1400 1600 1800 2000 Discharge (m3/s)

Fig. 49. Plot showing NdSN/YbSN ratios against discharge for the Rhine River.

Elderfield et al. (1990) report REE patterns for surface waters draining granites that are similar to those draining carbonates, and argue that aquatic chemistry dictates REE patterns in surface waters rather than REE patterns of the source rock. The influence of colloids is included in aquatic chemistry and plays an important role. Data presented here suggests that catchment geology plays at least a partial role in dissolved REE distributions, although catchment geology seems to affect major ion chemistry much more visibly. While discharge affects REE concentrations (higher discharge, higher REE), changes in REE distribution are not in full agreement with prediction. While an increase in pH is accompanied by a decrease in REE concentrations, there is also increased scatter at low pH (October 2009). Hence, the

114 controls on the natural REE distribution in the Rhine River are multifaceted and not readily broken down into components.

Estimating export rates of dissolved REE both via the tributaries into the Rhine River and via the Rhine River into the North Sea is not a trivial task. Increased discharge via direct rainfall could for instance cause dilution of REE in the surface waters, while runoff is likely to bring in significant amounts of colloids with it increasing REE concentrations. Additionally, the input of anthropogenic REE is not easily monitored or accounted for. The input of anthropogenic Gd depends on the number of MRI scans in the catchment of a WWTP and on the ratio of Gd‐ contaminated domestic sewage water relative to uncontaminated industrial sewage and surface run‐off. The input of La directly depends on the output from the catalyst production effluent. Both of these anthropogenic REE are also affected by the addition of waters (runoff, tributaries, etc.) that lack anthropogenic REE. Increased rainfall might also affect concentrations because of dilution of the anthropogenic signal, and/or increasing the (colloidal) particulate load of a river. Though not likely significant, loss of anthropogenic REE within the river due to sorption onto settling particles or even bioaccumulation is not ruled out in general. In the case of anthropogenic Gd, the environmental half‐life of Gd‐based contrast agents is rather high (Möller and Dulski, 2010a; 2010b; Möller et al., 2011) and these compounds are stable enough to be detected in the North Sea for example (Kulaksız and Bau, 2007). Additionally, significant Gd anomalies are found in tap water in Berlin due to natural and artificial groundwater recharge (Kulaksız and Bau, 2011a). The persistence of these Gd‐ based contrast agents through WWTP, groundwater infiltration and water treatment processes is also a good indicator of the long environmental half‐life of anthropogenic Gd in nature.

Consequently, estimations of the export of anthropogenic and natural REE vary greatly between sampling campaigns, and should not be taken as definite numbers but more as ranges. The main controlling factor here is the discharge data for each sampling date.

At our most downstream location sampled in all campaigns (Rhine‐km 811.4), anthropogenic

Gd (Gdanthr) concentrations are 7.67, 10.6 and 25.0 ng/kg on top of geogenic (natural) Gd

(Gdnat) concentrations of 3.61, 1.52 and 1.60 ng/kg in the respective samples of May 2008,

115 May 2009 and Oct 2009. Using these values together with discharge in the Rhine River at each sampling date, 53 kg (Oct 2009) to 215 kg (May 2008) of natural Gd export is estimated each year via the Rhine River towards the North Sea. Moreover, 457 kg (May 2008) to 828 kg (Oct 2009) annual anthropogenic Gd export is estimated via the Rhine River into the North Sea. While there are no samples presented here within the estuary of the Rhine River, the persistence of anthropogenic Gd through estuarine processes and its presence in coastal seawater has been shown elsewhere (Nozaki et al., 2000; Kulaksız and Bau, 2007).

Anthropogenic La concentrations (Laanthr) are 18.3, 19.6 and 26.0 ng/kg on top of natural La concentrations (Lanat) of 3.88, 1.77 and 2.14 ng/kg in the samples from May 2008, May 2009 and Oct 2009 respectively. Based on these numbers, estimates of anthropogenic La export via the Rhine River ranges between 862 kg (Oct 2009) and 1090 kg (May 2008) each year, together with export of 71 kg to 231 kg of natural La.

116 5. CONCLUSIONS AND OUTLOOK

This study represents the first detailed investigation of natural and anthropogenic REE in the Rhine River to our knowledge. The Rhine River shows rather high pH and low REE concentrations that increase with discharge. Shale‐normalized dissolved REE patterns are rather steep with LREE/HREE ratios generally lower than those reported for rivers elsewhere, but in agreement with previously published data for the Rhine River. Cerium anomalies increase with increasing REE concentrations and decreasing pH in the Rhine River. The variation in concentrations and distribution of dissolved REE can be partially explained by catchment geology. The shales of the Middle Rhine area cause an increase of shale normalized Nd/Yb ratios (NdSN/YbSN) in the Rhine River and its tributaries, due to the addition of more shale‐like components (NdSN/YbSN close to 1) in the REE content of the rivers. Discharge plays an important role, influencing dissolved REE distribution together with catchment geology. While variation in dissolved REE concentrations between sampling campaigns is relatively subtle, there is a clear increase of concentrations with increasing discharge within each sampling campaign. Additionally, NdSN/YbSN ratios increase with discharge at each location, likely due to the increased presence of colloids.

Catchment geology also plays an important role in controlling major ion chemistry. While most of the variation in both the Rhine River and its tributaries can be explained by catchment geology, there is also anthropogenic pressure for ions such as Na+ and Cl‐, although to a lesser extent than previously reported. The tributaries in the High Rhine and Upper Rhine partly influence the major ion chemistry of the Rhine River.

5.1 RARE EARTH ELEMENTS AS EMERGING CONTAMINANTS

The Rhine River is severely contaminated with anthropogenic Gd from diffuse sources (WWTP effluents) and anthropogenic La from a point source (effluent of a FCC catalyst producing factory). Anthropogenic Gd contamination can be observed throughout the Rhine River as well as its tributaries, while anthropogenic La is only observed in the Rhine River downstream of Rhine‐km 447.3. Discharge plays a diluting role for both anthropogenic Gd and La at each sampling location, and highest anomalies are found in the low discharge season. With anthropogenic Gd concentrations of up to 62‐times background concentrations,

117 anthropogenic Gd is a substantial part of the total REE budget of the Rhine River. Similarly, anthropogenic La concentrations downstream from Rhine‐km 447.3 can be as high as 46‐ times background concentrations. Both elements taken together can be as high as 90% of total REE concentrations in the Rhine River. Estimated yearly exports of anthropogenic Gd via the Rhine River into the North Sea range between 53 kg to 215 kg. Similarly each year, an estimated 862 to 1090 kg of anthropogenic La is exported into the North Sea, released from a point‐source. Future work should also include sampling the Rhine River estuary in order to confirm anthropogenic Gd input into the North Sea via the Rhine River system, and to check for the presence of La anomalies.

With the increasing applications of REE and demand for REE in the world, it is only a matter of time before a new anthropogenic REE is reported. These so‐called emerging contaminants are not frequently studied, and their effect on the ecosystem is principally unknown. Toxicological studies readily ignore low‐dosage low‐term and synergistic effects and instead focus on acute and chronic toxicity tests. While this information is essential, it is not well‐ suited for the profile of emerging contaminants.

This study will hopefully serve as a cautionary note for policy makers, with the message that emerging contaminants such as anthropogenic REE have already appeared in the hydrosphere, and will likely increase in the near future. If the current situation regarding lack of legal limits for REE in discharge effluents and responsibility for removing excess REE from effluent persists, in addition to the potential ecotoxicological effects of REE, geochemical studies of natural REE will also be hampered, possibly irreversibly.

5.2 SUGGESTIONS AND OUTLOOK

The interpretation of the dataset for natural REE suffers from lack of data that could contribute to the discussion. For example, the expected increase in LREE / HREE ratios with increasing REE concentrations is only partially observed in the Rhine River. Similarly Ce anomalies in literature decrease with increasing REE concentrations while this study suggests otherwise. DOC and Fe concentration data should reveal more information on the colloidal aspect of REE distribution. Without DOC or Fe data, it is not possible to comment on the

118 influence of colloids in detail, as the presence of colloids is only implied through discharge data in this study. Ultrafiltration alongside the routine 0.2 µm filtration should also reveal more information on the nature of colloids present and their effect on the distribution of REE in the Rhine River.

119 ACKNOWLEDGEMENTS

We thank Gila Merschel for her help with sampling, Jule Mawick and Daniela Meissner, Jacobs University Bremen, for their help in the Geochemistry Laboratory. Special thanks to Wilfried Wiechmann, Bundesanstalt für Gewässerkunde, Koblenz, for providing discharge data. This study was performed within the framework of DFG grant BA 2289/2‐1 to M.B.

120 6. REFERENCES

Bau, M., 1996. Controls on the fractionation of isovalent trace elements in magmatic and aqueous systems: evidence from Y/Ho, Zr/Hf, and lanthanide tetrad effect. Contributions to Mineralogy and Petrology V123, 323‐333.

Bau, M., Dulski, P., 1996. Anthropogenic origin of positive gadolinium anomalies in river waters. Earth and Planetary Science Letters 143, 245‐255.

Bau, M., Möller, P., Dulski, P., 1997. Yttrium and lanthanides in eastern Mediterranean seawater and their fractionation during redox‐cycling. Marine Chemistry 56, 123‐131.

Bau, M., Knappe, A., Dulski, P., 2006. Anthropogenic gadolinium as a micropollutant in river waters in Pennsylvania and in Lake Erie, northeastern United States. Chemie der Erde ‐ Geochemistry 66, 143‐152.

Berner, E.K., Berner, R.A., 1996. Global environment : water, air, and geochemical cycles. Prentice Hall, Upper Saddle River, NJ 376pp.

Buhl, D., Neuser, R.D., Richter, D.K., Riedel, D., Roberts, B., Strauss, H., Veizer, J., 1991. Nature and Nurture ‐ Environmental Isotope Story of the River Rhine. Naturwissenschaften 78, 337‐ 346.

Camusso, M., Vignati, D., van de Guchte, C., 2000. Ecotoxicological assessment in the rivers Rhine (The Netherlands) and Po (Italy). Aquatic Ecosystem Health & Management 3, 335‐345.

De Ruyter Van Steveninck, E., van Zanten, B., Admiraal, W., 1990. Phases in the development of riverine plankton: Examples from the rivers Rhine and Meuse. Aquatic Ecology 24, 47‐55.

Dupré, B., Viers, J., Dandurand, J.‐L., Polve, M., Bénézeth, P., Vervier, P., Braun, J.‐J., 1999. Major and trace elements associated with colloids in organic‐rich river waters: ultrafiltration of natural and spiked solutions. Chemical Geology 160, 63‐80.

Elderfield, H., Upstill‐Goddard, R., Sholkovitz, E.R., 1990. The rare earth elements in rivers, estuaries, and coastal seas and their significance to the composition of ocean waters. Geochimica et Cosmochimica Acta 54, 971‐991.

First Assessment of Transboundary Rivers, Lakes and Groundwaters, 2007. United Nations. Available online at http://www.unece.org/env/water/publications/pub76.html. Last accessed on May 25, 2012.

121 Gaillardet, J., Viers, J., Dupré, B., 2003. Trace Elements in River Waters, in: Heinrich, D.H., Karl, K.T. (Eds.), Treatise on Geochemistry. Pergamon, Oxford, pp. 225‐272.

Gibbs, R.J., 1970. Mechanisms Controlling World Water Chemistry. Science 170, 1088‐1090.

Hartmann, J., Jansen, N., Kempe, S., H, D., Uuml, Rr, H., 2007. Geochemistry of the River Rhine and the Upper Danube: Recent Trends and Lithological Influence on Baselines. Journal of Environmental Science for Sustainable Society 1, 39‐46.

Iliev, M. (2010). Effects of waste water treatment plant effluent on anthropogenic gadolinium concentrations in the Wupper River system. Unpublished B.Sc. thesis, Jacobs University Bremen, 36 p.

Ingri, J., Widerlund, A., Land, M., Gustafsson, O., Andersson, P., Ohlander, B., 2000. Temporal variations in the fractionation of the rare earth elements in a boreal river; the role of colloidal particles. Chemical Geology 166, 23‐45.

Johannesson, K.H., Lyons, W.B., Fee, J.H., Gaudette, H.E., McArthur, J.M., 1994. Geochemical processes affecting the acidic groundwaters of Lake Gilmore, Yilgarn Block, Western Australia: a preliminary study using neodymium, samarium, and dysprosium. J Hydrol 154, 271‐289.

Kempe, S., Krahe, P., 2005. Water and biogeochemical fluxes in the River Rhine catchment. Erdkunde 59, 216‐250.

Knappe, A., Möller, P., Dulski, P., Pekdeger, A., 2005. Positive gadolinium anomaly in surface water and ground water of the urban area Berlin, Germany. Chemie der Erde ‐ Geochemistry 65, 167‐189.

Kulaksız, S., Bau, M., 2007. Contrasting behaviour of anthropogenic gadolinium and natural rare earth elements in estuaries and the gadolinium input into the North Sea. Earth and Planetary Science Letters 260, 361‐371.

Kulaksız, S., Bau, M., 2011a. Anthropogenic gadolinium as a microcontaminant in tap water used as drinking water in urban areas and megacities. Applied Geochemistry 26, 1877‐1885.

Kulaksız, S., Bau, M., 2011b. Rare earth elements in the Rhine River, Germany: First case of anthropogenic lanthanum as a dissolved microcontaminant in the hydrosphere. Environ Int 37, 973‐979.

122 Kümmerer, K., Helmers, E., 2000. Hospital effluents as a source of gadolinium in the aquatic environment. Environmental Science & Technology 34, 573‐577.

Lawrence, M.G., Ort, C., Keller, J., 2009. Detection of anthropogenic gadolinium in treated wastewater in South East Queensland, Australia. Water Res 43, 3534‐3540.

Lawrence, M.G., Keller, J., Poussade, Y., 2010. Removal of magnetic resonance imaging contrast agents through advanced water treatment plants. Water Sci Technol 61, 685‐692.

Lee, J.H., Byrne, R.H., 1992. Examination of comparative rare earth element complexation behavior using linear free‐energy relationships. Geochimica et Cosmochimica Acta 56, 1127‐ 1137.

Meybeck, M., 2004. Global Occurrence of Major Elements in Rivers. In: Treatise on Geochemistry. Holland, H.D. and Turekian, K.K. (Editors), Elsevier, Amsterdam. 5: 207‐223.

Milich, L., Varady, R.G., 1999. Openness, Sustainability, and Public Participation: New Designs for Transboundary River Basin Institutions. The Journal of Environment & Development 8, 258‐306.

Moermond, C.T.A., Tijink, J., van Wezel, A.P., Koelmans, A.A., 2001. Distribution, speciation, and bioavailability of lanthanides in the Rhine‐Meuse estuary, the Netherlands. Environ Toxicol Chem 20, 1916‐1926.

Möller, P., Bau, M., 1993. Rare‐earth patterns with positive cerium anomaly in alkaline waters from Lake Van, Turkey. Earth and Planetary Science Letters 117, 671‐676.

Möller, P., Dulski, P., Bau, M., Knappe, A., Pekdeger, A., Sommer‐von Jarmersted, C., 2000. Anthropogenic gadolinium as a conservative tracer in hydrology. J Geochem Explor 69‐70, 409‐414.

Möller, P., Paces, T., Dulski, P., Morteani, G., 2002. Anthropogenic Gd in surface water, drainage system, and the water supply of the city of Prague, Czech Republic. Environmental Science & Technology 36, 2387‐2394.

Möller, P., Dulski, P., 2010a. Transmetallation of Gd‐DTPA by Cu, Y and lanthanides and its impact on the hydrosphere. Applied Geochemistry 25, 48‐59.

Möller, P., Dulski, P., 2010b. Gd‐DTPA in the hydrosphere: Kinetics of transmetallation by ions of rare earth elements, Y and Cu. Chem Erde‐Geochem 70, 125‐136.

123 Moller, P., Knappe, A., Dulski, P., Pekdeger, A., 2011. Behavior of Gd‐DTPA in simulated bank filtration. Applied Geochemistry 26, 140‐149.

Petelet‐Giraud, E., Klaver, G., Negrel, P., 2009. Natural versus anthropogenic sources in the surface‐ and groundwater dissolved load of the Dommel river (Meuse basin): Constraints by boron and strontium isotopes and gadolinium anomaly. J Hydrol 369, 336‐349.

Perret, D., Newman, M.E., Negre, J.C., Chen, Y.W., Buffle, J., 1994. Submicron Particles in the Rhine River .1. Physicochemical Characterization. Water Res 28, 91‐106.

Rabiet, M., Brissaud, F., Seidel, J.‐L., Pistre, S., Elbaz‐Poulichet, F., 2005. Deciphering the presence of wastewater in a medium‐sized Mediterranean catchment using a multitracer approach. Applied Geochemistry 20, 1587‐1596.

Rabiet, M., Togola, A., Brissaud, F., Seidel, J.L., Budzinski, H., Elbaz‐Poulichet, F., 2006. Consequences of treated water recycling as regards pharmaceuticals and drugs in surface and ground waters of a medium‐sized Mediterranean catchment. Environmental Science & Technology 40, 5282‐5288.

Rabiet, M., Brissaud, F., Seidel, J.L., Pistre, S., Elbaz‐Poulichet, F., 2009. Positive gadolinium anomalies in wastewater treatment plant effluents and aquatic environment in the Herault watershed (South France). Chemosphere 75, 1057‐1064.

Schijf, J., Debaar, H.J.W., Wijbrans, J.R., Landing, W.M., 1991. Dissolved Rare‐Earth Elements in the Black‐Sea. Deep‐Sea Res 38, S805‐S823.

Schijf, J., Debaar, H.J.W., Millero, F.J., 1995. Vertical Distributions and Speciation of Dissolved Rare‐Earth Elements in the Anoxic Brines of Bannock Basin, Eastern Mediterranean‐Sea. Geochimica et Cosmochimica Acta 59, 3285‐3299.

Shabani, M.B., Akagi, T., Masuda, A., 1992. Preconcentration of trace rare‐earth elements in seawater by complexation with bis(2‐ethylhexyl) hydrogen phosphate and 2‐ethylhexyl dihydrogen phosphate adsorbed on a C18 cartridge and determination by inductively coupled plasma mass spectrometry. Analytical Chemistry 64, 737‐743.

Shiller, A.M., 2002. Seasonality of dissolved rare earth elements in the lower Mississippi River. Geochem Geophy Geosy 3.

Sholkovitz, E.R., 1978. The flocculation of dissolved Fe, Mn, Al, Cu, Ni, Co and Cd during estuarine mixing. Earth and Planetary Science Letters 41, 77‐86.

124 Sholkovitz, E.R., Boyle, E.A., Price, N.B., 1978. The removal of dissolved humic acids and iron during estuarine mixing. Earth and Planetary Science Letters 40, 130‐136.

Sholkovitz, E.R., 1992. Chemical evolution of rare earth elements: fractionation between colloidal and solution phases of filtered river water. Earth and Planetary Science Letters 114, 77‐84.

Sholkovitz, E.R., 1993. The geochemistry of rare earth elements in the Amazon River estuary. Geochimica et Cosmochimica Acta 57, 2181‐2190.

Sholkovitz, E.R., 1995. The aquatic chemistry of rare earth elements in rivers and estuaries. Aquatic Geochemistry V1, 1‐34.

Singh, A.K., Hasnain, S.I., 1999. Environmental geochemistry of Damodar River basin, east coast of India. Environmental Geology 37, 124‐136.

Stogbauer, A., Strauss, H., Arndt, J., Marek, V., Einsiedl, F., van Geldern, R., 2008. Rivers of North‐Rhine Westphalia revisited: Tracing changes in river chemistry. Applied Geochemistry 23, 3290‐3304.

Tang, J., Johannesson, K.H., 2003. Speciation of rare earth elements in natural terrestrial waters: assessing the role of dissolved organic matter from the modeling approach. Geochimica et Cosmochimica Acta 67, 2321‐2339.

Tricca, A., Stille, P., Steinmann, M., Kiefel, B., Samuel, J., Eikenberg, J., 1999. Rare earth elements and Sr and Nd isotopic compositions of dissolved and suspended loads from small river systems in the Vosges mountains (France), the river Rhine and groundwater. Chemical Geology 160, 139‐158.

Van Nieuwenhuyse, E.E., 2007. Response of summer chlorophyll concentration to reduced total phosphorus concentration in the Rhine River (Netherlands) and the Sacramento – San Joaquin Delta (California, USA). Canadian Journal of Fisheries and Aquatic Sciences 64, 1529‐ 1542.

Verplanck, P.L., Taylor, H.E., Nordstrom, D.K., Barber, L.B., 2005. Aqueous stability of gadolinium in surface waters receiving sewage treatment plant effluent, Boulder Creek, Colorado. Environmental Science & Technology 39, 6923‐6929.

125 Verplanck, P.L., Furlong, E.T., Gray, J.L., Phillips, P.J., Wolf, R.E., Esposito, K., 2010. Evaluating the Behavior of Gadolinium and Other Rare Earth Elements through Large Metropolitan Sewage Treatment Plants. Environmental Science & Technology 44, 3876‐3882.

Wegman, R.C.C., Greve, P.A., De Heer, H., Hamaker, P., 1981. Methyl bromide and bromide‐ ion in drainage water after leaching of glasshouse soils. Water, Air, & Soil Pollution 16, 3‐11.

Weil, H., Knepper, T., 2006. Pharmaceuticals in the River Rhine, in: Knepper, T. (Ed.). Springer Berlin / Heidelberg, pp. 177‐184.

Wood, S.A., 1990. The aqueous geochemistry of the rare‐earth elements and yttrium : 1. Review of available low‐temperature data for inorganic complexes and the inorganic REE speciation of natural waters. Chemical Geology 82, 159‐186.

126 Online Resources:

ARA‐Röti, 2012. http://www.abfall‐sh.ch/roeti/index.php

German Federal Institute of Hydrology, 2012. www.bfg.de

ICPR. International catchment area of the Rhine: Properties, assessment of environ‐ mental effects of human activities and economic analysis of water use (part A). “Internationaal stroomgebiedsdistrict Rijn: Kenmerken, beoordeling van de milieueffecten van menselijke activiteiten en economische analyse van het watergebruik (deel A)”. Den Haag, the Netherlands: ICBR; 2005. p. 83. Available online at http://www.iksr.org/fileadmin/user_upload/Dokumente_nl/Rijnkaart/cc_02‐ 05nl_rev._18.03.05_online.pdf

International Association of Water Works in the Rhine Basin, 2012: http://www.iawr.org

KLIWA, 2012. http://www.kliwa.de/download/KLIWAHeft11.pdf

OneGeology Europe ‐ http://onegeology‐europe.eu/

Zweckverband Bodensee‐Wasserversorgung, 2012. www.zvbwv.de

127

128 CHAPTER V – ANTHROPOGENIC DISSOLVED AND

COLLOID/NANOPARTICLE‐BOUND SAMARIUM, LANTHANUM AND

GADOLINIUM IN THE RHINE RIVER AND THE IMPENDING

DESTRUCTION OF THE NATURAL RARE EARTH ELEMENT

DISTRIBUTION IN RIVERS

This chapter has been published in Vol. 362 of the journal Earth and Planetary Science Letters in 2013 and can be found online at http://dx.doi.org/10.1016/j.epsl.2012.11.033.

129 Earth and Planetary Science Letters 362 (2013) 43–50

Contents lists available at SciVerse ScienceDirect

Earth and Planetary Science Letters

journal homepage: www.elsevier.com/locate/epsl

Anthropogenic dissolved and colloid/nanoparticle-bound samarium, lanthanum and gadolinium in the Rhine River and the impending destruction of the natural rare earth element distribution in rivers

Serkan Kulaksız, Michael Bau n

Earth and Space Science Program, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany article info abstract

Article history: The strong increase in the consumption of rare earth elements (REE) in high-tech products and processes is Received 13 April 2012 accompanied by increasing amounts of REE released into the environment. Following the first report of Gd Received in revised form contamination of the hydrosphere in 1996, anthropogenic Gd originating from contrast agents has now been 1 November 2012 reported worldwide from river and estuarine waters, coastal seawater, groundwater and tap water. Recently, Accepted 20 November 2012 microcontamination with La, that is derived from a point source where catalysts for petroleum refining are Editor: G. Henderson Available online 23 December 2012 produced, has been detected in the Rhine River in Germany and the Netherlands. Here we report the occurrence of yet another REE microcontamination of river water: in addition to anthropogenic Gd and La, the Keywords: Rhine River now also shows significant amounts of anthropogenic Sm. The anthropogenic Sm, which enters anthropogenic rare earth elements the Rhine River north of Worms, Germany, with the same industrial wastewater that carries the samarium anthropogenicLa,canbetracedthroughtheMiddleand Lower Rhine to the Netherlands. At Leverkusen, lanthanum gadolinium Germany, some 250 km downstream from the point source at Worms, anthropogenic Sm still contributes up Rhine River ultrafiltration to 87% of the total dissolved Sm concentration of the Rhine River. Results from ultrafiltration suggest that nanoparticle while the anthropogenic Gd is not particle-reactive and hence exclusively present in the truly dissolved REE pool (o10 kDa), the anthropogenic La and Sm are also present in the colloidal/nanoparticulate REE pool (between 10 kDa and 0.2 mm). Though difficult to quantify, our data suggest that the Rhine River may carry up to 5700 kg of anthropogenic La, up to 584 kg of anthropogenic Sm, and up to 730 kg of anthropogenic Gd per year towardP the North Sea. There exist no regulatory limits for dissolved REE in natural waters, but total REE and Y ( REY) concentrations of up to 0.14 mg/kg in the plume downstream of and 52.2 mg/kg at the head of an effluent pipe at Rhine-km 447.3 at Worms get close to and well-above, respectively, the levels at which ecotoxicological effects have been documented. Because of the increasing use of REE and other formerly ‘‘exotic’’ trace elements in high-tech applications, these critical metals have now become emerging contaminants that should be monitored, and it appears that studies of their biogeochemical behavior in natural freshwaters might soon no longer be possible. & 2012 Elsevier B.V. All rights reserved.

1. Introduction Germany. The ultimate source of the elevated Gd concentrations are Gd-based contrast agents used in medical diagnostics (magnetic The rare earth elements (REE) belong to the group of critical metals resonance imaging, MRI), which reach surface waters with the clear- that are of strategic importance for the development of key technol- water discharge from wastewater treatment plants (WWTP). Since ogies, such as wind turbines, electrical car engines, medical diagnostics then, anthropogenic Gd has been reported worldwide, from rivers, and petroleum refining. World demand for REE is projected to increase lakes, ground water, tap water and coastal seawater (Bau and Dulski, from 136,000 t per year in 2010 to at least 185,000 t by 2015 1996; Bau et al., 2006; Elbaz-Poulichet et al., 2002; Knappe et al., (Humphries, 2011), and the rising consumption of REE for high-tech 2005; Kulaksız and Bau, 2007, 2011a, 2011b; Lawrence et al., 2006; products and processes leads to the release of increasing amounts of Lawrence, 2010; Moller¨ et al., 2000, 2002, 2003; Morteani et al., 2006; REE into the environment, either as solid or as dissolved phase. Nozaki et al., 2000; Petelet-Giraud et al., 2009; Rabiet et al., 2005, The first report of an anthropogenic REE component in natural 2009; Tricca et al., 1999; Verplanck et al., 2005; Zhu et al., 2004, 2005). waters was published in the mid-1990s (Bau and Dulski, 1996)when Recently, we have shown that the Rhine River, Germany, carries anomalously high concentrations of Gd were detected in rivers in significant amounts of anthropogenic La and to a lesser extent other light REE (LREE), apparently as a dissolved (o0.2 mm-sized) microcontaminant (Kulaksız and Bau, 2011a). This La contamina-

n tion has also been observed in the Rhine River in the Netherlands Corresponding author. Tel.: þ49 421 2003564. E-mail address: [email protected] (M. Bau). (Verheul et al., 2011). We had suggested that anthropogenic

0012-821X/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2012.11.033

130 S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50

5°E 6°E 7°E 8°E 9°E

0150 00010.5 km km 52 ° N

Leerdam Lippe

NL Essen Dortmund Rhine-km 448.9 Duisburg Ruhr Düsseldorf Monheim Wupper 51 ° N Leverkusen Rhine-km 447.3 Köln Sieg Lahn Rhine-km 446.7

Erft B Rhine River Ahr Pfrimm Creek Frankfurt am Main 50 ° N

L Main Nahe Worms D ! Worms 49 ° N Mosel

F Stuttgart Sample Sites Strasbourg Large City Neckar N Ill * SmSN / Sm SN: 200 48 ° N

Fig. 1. Simplified map of (A) the Middle and Lower Rhine River and (B) a close-up of the area north of Worms, Germany, showing sampling sites and the location where the n anthropogenic Sm anomaly (illustrated as SmSN/SmSN) first occurs. For further explanation see text.

contamination of the hydrosphere with formerly ‘‘exotic’’ trace 946), the Netherlands (Fig. 1a). To check for potential analytical elements would likely accelerate in the near future and, using the artifacts, we also sampled and processed water from the pristine REE as examples, had cautioned that these emerging microconta- Wiembach Creek at Leverkusen–Opladen, Germany. In spring of minants would likely soon present very severe difficulties for 2012 (1 April, ’12), we sampled the Rhine River between Rhine-km studies of the biogeochemical behavior of these elements in 446.7 and 448.9 (Fig. 1b), i.e. 600 m upstream and 1600 m down- pristine environments. stream from the site where industrial effluent enters the river at Only a few months later, this pessimistic view is already Rhine-km 447.3 and causes the previously observed contamination confirmed. We here report another anthropogenic REE micro- with La (Kulaksız and Bau, 2011a). Unfortunately, the head of the contamination: between mid-October 2010 and mid-May 2011 effluent pipe was inaccessible due to high water level of the Rhine small but significant amounts of anthropogenic Sm have started River and the sample closest to the pipe could only be taken at about to appear in the Rhine River in Germany and the Netherlands. 2 m distance downstream and was already substantially diluted Ultrafiltration revealed that, similar to anthropogenic La and with Rhine River water. Sampling, sample treatment, and chemical Gd, part of this anthropogenic Sm is truly dissolved. However, analyses followed our routine protocol as described previously in contrast to the anthropogenic Gd that is exclusively present (Kulaksız and Bau, 2007, 2011a, 2011b; and references therein), in the truly dissolved REE pool, considerable fractions of with the exception of the three samples from 1 April, ’12, which anthropogenic Sm and La are also bound to colloids and were analyzed without any preconcentration. nanoparticles present in the river water. The good overall reproducibility of the method employed is shown by the very close match of measured REE concentrations of two independently processed aliquots of a single Rhine River sample 2. Methods from 14 May, ’11 (Table 1). We also emphasize that the pristine Wiembach Creek does not show any anthropogenic REE anomalies 2.1. Sampling and analysis (Fig. 2), demonstrating that the anomalies observed in the Rhine River samples do not represent sampling or analytical artifacts. Surface water samples from the Rhine River were collected at In addition to the 0.2 mm filtration, one sample from the Leverkusen (Rhine-km 703), Germany, and at Leerdam (Rhine-km Rhine River (18 January, ’12) was additionally ultrafiltered

131 Table 1 n n n Sample characterization and REE concentrations (ng/kg) for the Rhine River. La ,Sm ,Gd and all related numbers calculated using Eqs. (1) and (2) .

Wiembach Rhine River Rhine Rhine River Rhine Rhine River Rhine Rhine Rhine River Rhine River Rhine River Rhine Rhine Effluent Effluent Creek River (duplicate) River River River River River Plume

Leverkusen– Monheim Leverkusen Leverkusen Leverkusen Leverkusen Leverkusen Leerdam Leverkusen Leverkusen Leverkusen Worms Worms Worms Worms Opladen

19 January ’12 16 October 14 May ’11 14 May ’11 26 August 7 November 1 January 2 January 18 January 18 January 18 January 31 March 31 March 31 March 27 October ’10 ’11 ’11 ’12 ’12 ’12 ’12 ’12 ’12 ’12 ’12 ’10 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 10 kDa Nanopartic. o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 mm o 0.2 Latitude 51.0737 51.11604 51.04346 51.04346 51.04346 51.04346 51.04346 51.827994 51.04346 51.04346 51.04346 49.66064 49.67845 49.66499 49.66499 Longitude 7.00098 6.87915 6.94483 6.94483 6.94483 6.94483 6.94483 5.09054 6.94483 6.94483 6.94483 8.36607 8.35346 8.36176 8.36176 Rhine-km – 715.7 703 703 703 703 703 946.2 703 703 703 446.7 448.9 447.3 447.3 pH 7.51 7.95 8.53 8.53 7.55 7.84 7.78 7.25 8.17 8.21 – – – – 6.95 .Klkı,M a at n lntr cec etr 6 21)43–50 (2013) 362 Letters Science Planetary and Earth / Bau M. Kulaksız, S. Cond. 142 855 642 642 514 705 524 – 544 603 – – – – 26,750 (ms/cm) Y 170 17.3 9.18 9.22 12.3 12.9 37.5 38.5 33.8 17 16.9 – – 296 4610 La 89 104 12.6 13 29.4 36.1 78.7 51.2 71.3 18.9 52.4 31.5 7790 115,000 49,300,000 Ce 141 9.2 1.16 1.34 4.14 1.54 29 23.5 22.2 3.78 18.5 59.9 172 9340 2,550,000 Pr 28.8 1.56 0.516 0.538 0.986 0.989 5.35 4.72 4.22 1.04 3.18 7.82 87.3 1280 108,000 Nd 132 6.59 2.85 2.96 4.12 3.28 23.1 21.6 19 5.04 13.9 30.4 70.1 660 185,000 Sm 31.7 1.91 8.09 8.39 6.03 5.97 12.7 8.37 10.6 3.59 7.04 7.96 852 13,100 13,000 Eu 7.13 0.455 0.313 0.318 0.392 0.248 1.46 1.27 1.14 0.387 0.753 1.51 3.57 3.68 2010 Gd 34.8 21.6 13.8 14.1 10.4 16.4 11 13.2 8.68 5.3 3.38 34 34.9 44.8 3040 Tb 4.81 0.373 0.243 0.25 0.243 0.203 0.998 0.955 0.751 0.305 0.445 1.26 1.87 3.02 253 132 Dy 28.8 2.22 1.51 1.57 1.55 1.43 6.08 5.68 4.67 2.13 2.54 6.95 10.7 11.1 632 Ho 5.6 0.58 0.329 0.344 0.392 0.366 1.32 1.28 1.02 0.528 0.49 1.55 2.01 2.48 488 Er 15.4 1.85 1.09 1.16 1.43 1.46 3.92 4.06 3.12 1.81 1.32 4.22 5 8.45 227 Tm 2 0.296 0.185 0.198 0.228 0.262 0.547 0.552 0.447 0.283 0.161 0.703 0.915 1.68 17.5 Yb 12.7 2.34 1.55 1.67 1.86 2.4 3.78 3.82 3.18 2.2 0.975 6.78 8.77 13.9 166 PLu 1.88 0.425 0.318 0.345 0.407 0.687 0.657 0.677 0.58 0.389 0.191 1.31 2.01 2.95 42.5 PREE 536 153 44.6 46.2 61.6 71.4 178 141 151 45.6 105 196 9040 139,800 52,50,000 REY 706 171 53.8 55.4 73.9 84.4 216 179 185 62.6 122 – – 140,100 52,760,000 n 87 3.4 0.93 0.99 1.6 1.6 13 13 11 2.4 9.3 22 49 4300 – La n Sm 31 1.8 1.1 1.1 1.4 0.96 6 5.3 4.7 1.5 3.2 6.7 16 34 – Gdn 37 2.5 2 2.1 2.4 1.4 7.9 6.7 6.1 2.2 3.9 7.6 18 8.9 – Laanthr 1.6 100 12 12 28 35 66 38 60 17 43 9.6 7700 110,000 – Smanthr 1 0.1 7 7.3 4.6 5 6.7 3.1 5.9 2.1 3.8 1.3 840 13,000 – Gdanthr – 19 12 12 8 15 3 6.5 2.6 3.1 – –– – – n LaSN /LaSN 1 30 14 13 19 23 6 3.9 6.3 8 5.7 1.4 160 27 – n SmSN /SmSN 1 1.1 7.6 7.7 4.3 6.2 2.1 1.6 2.2 2.4 2.2 1.2 54 390 – n GdSN /GdSN 0.95 8.5 6.8 6.8 4.3 11 1.4 2 1.4 2.4 0.87 4.5 1.9 5.1 –

%Laanthr 1.8 97 93 92 95 96 83 74 84 88 82 30 99 96 – %Smanthr 3.2 5.4 87 87 77 84 53 37 55 59 54 16 98 100 – %Gdanthr – 888585779128493058– – – – –

Tm concentrations in italics are interpolated and not measured (see text). Sample ‘‘nanopartic.’’ is the difference between the o 0.2 mm and o 10 kDa fractions (see text). S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50 through a cross-flow ultrafiltration system. 3000 mL of unacidi- We report concentrations in the o0.2 mm REE pool as ‘‘dissolved’’ fied sample volume were passed through a Vivaflow 50 unit (i.e. comprised of REE bound to colloids/nanoparticles, truly dissolved (10 kDa MWCO) similar to that described by Schlosser and Croot REE and dissolved chemical REE complexes), while REE concentra- (2008), before the ultrafiltrate was processed using the same tions in the o10 kDa pool are referred to as ‘‘truly dissolved’’. The separation/preconcentration method employed in analyzing the difference between the truly dissolved and the dissolved REE con- other samples. centrations represents the colloid/nanoparticle-bound REE concentra- tion which we refer to as ‘‘nanoparticulate’’ REE pool for convenience.

2.2. Quantification of REE anomalies

In shale-normalized REE patterns (REESN, Post Archean Australian Shale, PAAS, McLennan, 1989) of pristine rivers, the elements from La to Gd form one coherent group, while Tb to Lu form another one. The two groups are separated by a small ‘‘step’’ down between Gd and Tb (Fig. 2, Wiembach Creek). Because of the numerous REE anomalies in Rhine River water the following equations have to be

used to quantify the anomalies in the REESN patterns: n log LaSN ¼ð2logNdSNlog EuSNÞð1aÞ

n log SmSN ¼ð2 log EuSN þlog NdSNÞ=3 ð1bÞ

n log GdSN ¼ ðÞ4logEuSNlog NdSN =3 ð1cÞ where the subscript SN denotes normalization to PAAS, and the superscript n denotes the geogenic background (extrapolated/ interpolated). Note that this approach cannot be applied to slightly acidic rivers rich in organic colloids, such as the Amazon River and other

tropical rivers, as these typically show middle REE-enriched REESN patterns and not the subdivision into a La–Gd and Tb–Lu trend observed in rivers in temperate latitudes. We emphasize that we use Eu to quantify and/or illustrate the anthropogenic Sm input, although Eu can be decoupled from the other REE due to its unique redox behavior and, therefore, may show anomalies in normalized REE patterns. However, because the major tributaries to the Rhine River and the main river itself upstream from Worms Fig. 2. REESN patterns of pristine Wiembach Creek at Leverkusen showing smooth do not show any EuSN anomaly, this element can be used in the REE patterns without any anthropogenic anomalies, the Rhine River at Leverkusen (Bau and Dulski, 1996) showing a small anthropogenic Gd anomaly, and the Rhine normalization procedure of these samples. We caution, however, River at Monheim (2010) showing large positive anthropogenic La and Gd that this prerequisite has to be verified before Eq. (1b) is applied anomalies. to a river water data set.

Fig. 3. REESN patterns of the Rhine River sampled during low-discharge periods in 2011 (A) and high-discharge periods in 2012 (B), showing anthropogenic positive La, Sm, and Gd anomalies.

133 S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50

For each anomalous REE, the anthropogenic component can be previously published data for the Rhine River (Kulaksız and Bau, calculated using the following equations: 2011a). Anthropogenic Sm concentrations at Leverkusen range between REEn ¼ REEn ½REE ð2aÞ SN PAAS 5.0 and 7.3 ng/kg on top of geogenic background concentrations of 0.96 and 1.1 ng/kg, respectively; the anthropogenic Sm represents n REEanthropogenic ¼ REEmeasuredREE ð2bÞ 84% and 87% of total Sm. North of Worms, around Rhine River km 447.3 (Fig. 1b) where where REE may be La, Sm or Gd, and [REE ] is the concentra- PAAS industrial effluents from a production facility for catalysts used in tion of the respective REE in PAAS. petroleum refining enter the river, the anthropogenic REE load Note that positive Sm anomalies which have been observed in changes substantially. From 600 m upstream of the effluent pipe international reference standards of seawater and river water, are to 1600 m downstream of it, Sm and La concentrations increase considered artifacts of standard preparation (e.g., Lawrence and strongly from 7.96 ng/kg to 852 ng/kg and from 31.5 ng/kg to Kamber, 2007; and references therein) or artifacts of sample 7790 ng/kg, respectively. At a distance of 2 m downstream from filtration (Moller¨ et al., 2003), and are not due to an anthropo- the head of the effluent pipe, i.e. within the plume where pure genic Sm contamination of the water. We reiterate that the lack of effluent mixes and is already substantially diluted with Rhine any Sm (and La and Gd) anomaly in samples from the pristine P River water, REE concentrations are highest ( REE¼139.8 mg/kg) Wiembach Creek (Fig. 2), that were processed together with the with very significant anomalous enrichments of La (115 mg/kg) Rhine River samples, shows that such artifacts did not affect and Sm (13.1 mg/kg). While the sample from 600 m upstream of our data. the effluent pipe shows only a strong positive Gd anomaly, the samples from 2 m and 1600 m downstream of the pipe head also display pronounced positive La and Sm anomalies (Fig. 4). 3. Results

3.1. Dissolved REE 3.2. The truly dissolved and nanoparticulate REE pools

Dissolved concentrations of purely geogenic Yb in the Rhine In order to investigate the physical speciation of the dissolved River samples taken at Leverkusen vary by a factor of 2.4 and anthropogenic La, Sm and Gd, i.e. their association with truly cover a rather small range from 1.55 ng/kg (14 May, ’11) to dissolved compounds and/or colloids/nanoparticles, an aliquot of 3.78 ng/kg (1 January, ’12). This concentration range is similar the Rhine River sample from 18 January, ’12, was subjected to to previously published REE data for the Rhine River at the same ultrafiltration (o10 kDa). Concentrations in the nanoparticulate location (3.88 ng/kg, Bau and Dulski, 1996; and 2.44 ng/kg, pool (10 kDa–0.2 mm) have been calculated by subtracting con- Kulaksız and Bau, 2011a), and to the general range observed in centrations in the truly dissolved pool (o10 kDa ultrafiltrate) other European rivers (Gaillardet et al., 2003). The Rhine River from those in the dissolved (o0.2 mm) pool: samples with highest REE concentrations have been taken during times of high discharge and turbidity in January 2012 (Table 1) REEnanoparticulate ¼ REE o 0:2 mmREE o 10 kDa ð3aÞ and the high concentrations are very likely related to the presence of large amounts of colloids and nanoparticles.

The REESN pattern of the Rhine River sample taken at Monheim (12.7 km downstream from Leverkusen) in October 2010 (Fig. 3) is characterized by prominent anthropogenic positive anomalies n n of La (LaSN/LaSN¼30) and Gd (GdSN/GdSN¼8.5). While the anthro- pogenic Gd is derived from diffuse sources (WWTP effluents in the catchment of the Rhine River and its tributaries), the anthro- pogenic La originates from a point source at Rhine-km 447.3, where catalysts for petroleum refining are produced. This parti- cular sample (Monheim, October ’10) carries no other anthropo- genic anomalies and confirms the findings of Kulaksız and Bau (2011a). In marked contrast, all Rhine River samples taken at Leverku- sen after October 2010 show anomalously high Sm concentrations (Fig. 3) that are up to 7.7 times higher than geogenic background values, producing a positive Sm anomaly in REESN patterns (SmSN/ n SmSN ¼2.17.7). These samples also show anthropogenic positive n n La (LaSN/LaSN¼6.023) and Gd anomalies (GdSN/GdSN¼1.411). Similar to the La and Gd anomalies, this positive Sm anomaly still persists at Leerdam (Rhine-km 946), the Netherlands, some 243 km downstream from Leverkusen (Fig. 2). We emphasize that our control sample from Wiembach Creek does not show any positive La, Gd or Sm anomaly (Fig. 2). Our post-October 2010 REE data from the Rhine River at Leverkusen, Germany, represents the first report of dissolved anthropogenic Sm anomalies found in natural waters. P Bulk dissolved REE ( REE) concentrations at Leverkusen range Fig. 4. REE patterns of samples taken 400 m upstream and 2 m and 1600 m from 45 ng/kg (14 May, ’11) to 179 ng/kg (1 January, ’12). SN downstream of an effluent pipe at Rhine-km 447.3. Note the high REE concentra- Anthropogenic La accounts for 483% of total La and anth- tions and the first appearance of positive La and Sm anomalies downstream of the ropogenic Gd for 477% of total Gd. This is consistent with effluent pipe.

134 S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50

speciation should differ from that of anthropogenic La. The spatial distribution of the anomalous Sm enrichment (Fig. 1b and Fig. 4) clearly demonstrates that the anthropogenic Sm originates from the same industrial cracking catalyst production effluent that causes the La contamination. Like other REE, Sm finds applications in many different areas, from high-strength permanent magnets to control rods in nuclear reactors. It is used as a catalyst in assisting the decomposition of plastics, the dechlorination of polychlorinated biphenyls (PCBs) (Emsley, 2011), and the dehydration and dehydrogenation of ethanol (Hammond, 2011); Sm(II) iodide is commonly used as a reducing and coupling agent in organic synthesis (Girard et al., 1980).

4.2. Truly dissolved and nanoparticulate anthropogenic REE

Comparison of REE concentrations (Table 1) in the dissolved and truly dissolved pools of Rhine River water (18 January, ’12) show that removal of geogenic REE by ultrafiltration is most significant for the LREE and decreases systematically with increasing atomic number from 83% for Ce and 66% for Eu down to only 30% for Yb and Lu. This is in line with previous results that have suggested that the heavy REE occur predominantly in the truly dissolved pool, whereas the light REE are more particle- reactive and are rather associated with the nanoparticulate pool (Elderfield et al., 1990; Sholkovitz, 1992, 1995).

Fig. 5. REESN patterns of the dissolved (o0.2 mm), nanoparticulate (410 kDa and The lack of a positive Gd anomaly in the nanoparticulate REE o0.2 mm) and truly dissolved (o10 kDa) REE pools in the Rhine River sampled at pool (Fig. 5) demonstrates that anthropogenic Gd is not particle- Leverkusen (18 January, ’12). Notice that in marked contrast to Gd, significant La reactive, and that it exclusively partitions into the truly dissolved and Sm anomalies are also present in the nanoparticulate pool. REE pool. This agrees with previous results (Knappe et al., 2005; Kulaksız and Bau, 2007; Kunnemeyer¨ et al., 2009; Morteani et al., 2006) suggesting that anthropogenic Gd speciation is dominated Removal of REE through ultrafiltration, i.e. the percentage by highly stable and water soluble chemical complexes with long associated with nanoparticulates, can be calculated by environmental half-lifes (e.g., Gd-based contrast agents, such as %removal ¼ REEnanoparticulate=REE o 0:2 mm 100 ð3bÞ Gd-DTPA). In marked contrast, La and Sm show no clear pre- P ference for the truly dissolved or the nanoparticulate pool (Fig. 5), Bulk REE concentration (151 ng/kg REE) in the dissolved pool demonstrating that anthropogenic La and Sm are both particle- ( 0.2 mm, 18 January, ’12) is at the high end of the concentration o reactive and behave rather similar to the geogenic REE. For La, this range of our dataset, but similar to concentrations in the water is in agreement with results of Verheul et al. (2011) for the Dutch sampled at Leerdam and Leverkusen in January ’12, during a time part of the Rhine River, that demonstrate that the anthropogenic of high discharge and turbidity. After ultrafiltration, the truly P positive La anomaly is also observed in the 40.45 mm sized dissolved pool ( 10 kDa, 18 January, ’12) carries 46 ng/kg REE, o particulate REE pool. with the difference of 105 ng/kg attributed to the nanoparticulate pool. Dissolved Yb in the dissolved, truly dissolved and nanopar- ticulate pools are 3.18, 2.20 and 0.975 ng/kg respectively. 4.3. Quantifying anthropogenic REE transport via the Rhine River

Fig. 5 shows the REESN patterns of the truly dissolved and nanoparticulate REE pools in Rhine River water (18 January, ’12). The amounts of anthropogenic REE transported via the Rhine While the truly dissolved pool displays anthropogenic positive La River toward the North Sea are difficult to quantify: input of n n n (LaSN/LaSN¼8.0), Sm (SmSN/SmSN¼2.4) and Gd (GdSN/GdSN ¼2.4) anthropogenic Gd depends on the number of MRI scans in the anomalies in a REESN pattern that closely resembles those of the catchment of a WWTP and on the ratio of Gd-contaminated Rhine River during times of low river discharge (Fig. 3a), the domestic sewage water relative to uncontaminated industrial n nanoparticulates only show positive La (LaSN/LaSN¼5.7) and Sm sewage and surface run-off, while input of La and Sm depends n (SmSN/SmSN ¼2.2) anomalies, but no positive Gd anomaly, in a on the level of catalyst production. There might be loss of general REESN pattern that is slightly enriched in the middle REE. anthropogenic REE within the river due to sorption onto settling particles or even bioaccumulation. Rainfall causing increasing river discharge might also affect concentrations because of dilu- 4. Discussion tion of the anthropogenic signal, and/or increasing the (colloidal) particulate load of a river. 4.1. Samarium as another emerging REE microcontaminant Hence, depending on the sampling date, anthropogenic REE transport by the Rhine River varies significantly; at Leverkusen, Significant amounts of anthropogenic Sm first appeared in the for example, between 0.9–15.5 kg/day for La, 0.4–1.6 kg/day for Rhine River at some time between October 2010 and May 2011 Sm, and 0.9–2.0 kg/day for Gd. Despite the large range, these (Fig. 3). The presence of both, anthropogenic Sm and La, in the numbers suggest that the Rhine River alone carries every year ultrafiltrate of Rhine River water (Fig. 5) indicates that, similar to between 330–5700 kg of anthropogenic La, 146–584 kg of anthro- anthropogenic La, anthropogenic Sm occurs as a water-soluble Sm pogenic Sm, and 329–730 kg of anthropogenic Gd toward the compound, and hence there is no obvious reason why its chemical North Sea. These amounts are not trivial, particularly considering

135 S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50 that the anthropogenic La and Sm appear to be derived from a facility producing catalysts for petroleum refining, that causes the single point source. La contamination in the Rhine River. Similar to anthropogenic La, but in marked contrast to anthropogenic Gd, this Sm originates 4.4. Environmental impact of anthropogenic REE from a point source. Anthropogenic dissolved Sm concentrations (in the o0.2 mm filtrate) are as high as 7.30 ng/kg and constitute The (eco)toxicity of the REE not only depends on their concen- up to 87% of the total dissolved Sm. Even at Leerdam (at Rhine-km tration but also on the mode of administration, with the intraper- 946, some 240 km downstream of Leverkusen and 500 km down- itoneal route resulting in much higher toxicity levels than the oral stream of the point source), 3.1 ng/kg of anthropogenic Sm is route (Bruce et al., 1963; Haley et al., 1961, 1963, 1964a, 1964b, observed, which represents 37% of the total dissolved Sm. All 1966). Since all REE have similar ionic radii and are mostly trivalent, anthropogenic elementsP (La, Sm, and Gd) taken together consti- they not only show similar geochemical behavior, but also similar tute up to 82% of REE in the Rhine River at Leverkusen. While (eco)toxicological behavior. The calcium inhibiting function of La3þ , these concentrations are well-below the threshold of about 1 mg/ for example, is at least qualitatively shared by other trivalent REE kg at which ecotoxicological effects have been observed, REY (Weiss and Goodmann, 1975). Hence, when discussing (eco)toxico- concentrations are significantly higher in the effluentP plume downstream of the effluent pipe at Rhine-km 447.3 ( REY up logicallyP effective concentrations, the concentration of bulk REE and Y( REY), i.e. total REE concentration including the concentration of to 52 mg/kg) and exceed the above mentioned ecotoxicological Y (which is the geochemical twin element of Ho and hence shows threshold by one and a half orders of magnitude. The addition of anthropogenic Sm anomalies to the already similar behavior, e.g., Bau, 1996), should beP discussed, rather than the concentration of an individual REE. The REY concentration in severely distorted REE pattern of the Rhine River signals once the truly dissolved, nanoparticulate and dissolved pools of Rhine again the immediate danger that is ahead both for the quality of River water at Leverkusen (18 January, ’12) are 185 ng/kg, 63 ng/kg, natural waters and for our ability to study them using REE. With the rapidly increasing demand for REE and the rising number of and 122 ng/kg, respectively. P Unfortunately, the bulk REY concentration ( REY) is usually applications, more anthropogenic REE are expected to enter the not considered and existing experimental studies have focused on environment in the near future. Hence, the fate and pathways of individual REE instead. Shi et al. (2006a, 2006b), for example, anthropogenic REE should be thoroughly studied, including any reported pathological effects in the kidneys and livers of rats potential (and synergetic) toxicity, while distribution and beha- subjected to Sm nitrate at Sm concentrations in the mg/kg range vior of the natural REE in different compartments of the environ- and demonstrated degradation in learning and memory functions ment should be closely monitored. The former is essential from a in populations exposed to low Sm (3.0 mg/kg) and intermediate health point of view, and the need for the latter is acute, as Sm (4.5 mg/kg) concentrations. Liver damage in rats occurs at establishing REE background concentrations in industrialized La concentrations of 20 mg/kg (Chen et al., 2003), and Zhang et al. countries will soon no longer be possible due to widespread (2010) have shown significant adverse effects on the growth and contamination. Considering that the increasing use of formerly reproduction of worms (Caenorhabditis elegans) at La concentra- exotic high-tech metals is not confined to the REE, the situation tions above 1.39 mg/kg. Moreover, Sun et al. (1996) have also regarding the rare earths is just the prelude of what may soon shown that bioaccumulation of LREE occurs in internal organs of happen with other trace elements that are destined to become emerging contaminants. carp (Cyprinus carpio L.). P Concentrations of dissolved Sm and REY in the lower reaches of the Rhine River at Leverkusen are below concentrations found to be toxic in these studies. However, the highest dissolved Acknowledgments Sm concentration in the effluent plume (sampled on 1 April, ’12) We thank Nathalie Tepe, Jule Mawick and Daniela Meissner, is 13.1 mg/kg at Rhine-km 447.3 and still 0.85 mg/kgP some 1600 m downstream at Rhine-km 448.9. The dissolved REY concentra- Jacobs University Bremen, for their help in the Geochemistry tion at Rhine-km 447.3 is 140.1 mg/kg. The purest effluent Laboratory, and Eike Breithaupt, Dunedin & Kiel, for his comments sampled to date (sampled during a low-discharge period on 27 on ultrafiltration using Vivaflow units. Special thanks to Wilfried October, ’10 (Kulaksız and Bau, 2011a), i.e. before the occurrence Wiechmann, Bundesanstalt fur¨ Gewasserkunde,¨ Koblenz, for providing discharge data. We appreciate the efforts of EPSL editor of an anthropogenicP Sm anomaly in the Rhine River) shows a dissolved REY concentration as high as 52.2 mg/kg (Table 1). G. Henderson and two anonymous reviewers. This study was per- Nevertheless, considering the exact sampling spot, even this formed within the framework of DFG grant No. BA 2289/2-1 to M.B. sample was diluted by ‘‘normal’’ Rhine River water, suggesting that the concentrations observed are belowP those of pure effluent. References We emphasize that the dissolved REY concentrations detected are up to 50 times above the concentration levels at Bau, M., 1996. Controls on the fractionation of isovalent trace elements in which ecotoxicological effects have been observed. Considering magmatic and aqueous systems: evidence from Y/Ho, Zr/Hf, and lanthanide that bioaccumulation of REE has been observed in carp (Sun et al., tetrad effect. Contrib. Mineral. Petrol. V123, 323–333. Bau, M., Dulski, P., 1996. Anthropogenic origin of positive gadolinium anomalies in 1996) and that the river bank close to the effluent pipe is a river waters. Earth Planet. Sci. Lett. 143, 245–255. popular fishing spot, a study focusing on the potential effects of Bau, M., Knappe, A., Dulski, P., 2006. Anthropogenic gadolinium as a micropollu- this REE (micro)contamination on the local fish population seems tant in river waters in Pennsylvania and in Lake Erie, northeastern United States. Chem. Erde 66, 143–152. warranted. Bruce, D.W., Hietbrink, B.E., Dubois, K.P., 1963. The acute mammalian toxicity of rare earth nitrates and oxides. Toxicol. Appl. Pharmacol 5, 750–759. Chen, D., Liu, Y., Chen, A.-J., Nie, Y.-X., 2003. Experimental study of subchronic 5. Conclusions toxicity of lanthanum nitrate on liver in rats. Nonlinearity Biol. Toxicol. Med. 1, 469–480. Elbaz-Poulichet, F., Seidel, J.-L., Othoniel, C., 2002. Occurrence of an anthropogenic This study of the REE distribution in the Rhine River, Germany, gadolinium anomaly in river and coastal waters of Southern France. Water is the first report of an anthropogenic Sm anomaly in natural Res. 36, 1102–1105. Elderfield, H., Upstill-Goddard, R., Sholkovitz, E.R., 1990. The rare earth elements in waters. The anthropogenic Sm enters the Rhine River at the same rivers, estuaries, and coastal seas and their significance to the composition of site at Worms, Germany, via the same industrial effluent from a ocean waters. Geochim. Cosmochim. Acta 54, 971–991.

136 S. Kulaksız, M. Bau / Earth and Planetary Science Letters 362 (2013) 43–50

Emsley, J., 2011. Nature’s Building Blocks: An A–Z Guide to the Elements, new ed. Moller,¨ P., Paces, T., Dulski, P., Morteani, G., 2002. Anthropogenic Gd in surface Oxford University Press, Oxford. water, drainage system, and the water supply of the city of Prague, Czech Gaillardet, J., Viers, J., Dupre´, B., 2003. Trace elements in river waters. In: Heinrich, Republic. Environ. Sci. Technol. 36, 2387–2394. D.H., Karl, K.T. (Eds.), Treatise on Geochemistry. Pergamon, Oxford, Morteani, G., Moller,¨ P., Fuganti, A., Paces, T., 2006. Input and fate of anthropogenic pp. 225–272. estrogens and gadolinium in surface water and sewage plants in the hydro- Girard, P., Namy, J.L., Kagan, H.B., 1980. Divalent lanthanide derivatives in organic logical basin of Prague (Czech Republic). Environ. Geochem. Health 28, synthesis. 1. Mild preparation of samarium iodide and ytterbium iodide and 257–264. their use as reducing or coupling agents. J. Am. Chem. Soc. 102, 2693–2698. Nozaki, Y., Lerche, D., Alibo, D.S., Tsutsumi, M., 2000. Dissolved indium and rare Haley, T.J., Efros, M., Komesu, N., Koste, L., Upham, H.C., 1964a. Pharmacology and earth elements in three Japanese rivers and Tokyo Bay: evidence for anthro- toxicology of lutetium chloride. J. Pharm. Sci. 53, 1186–1188. pogenic Gd and In. Geochim. Cosmochim. Acta 64, 3975–3982. Haley, T.J., Flesher, A.M., Komesu, N., Upham, H.C., Cawthorne, J., Mavis, L., 1963. Petelet-Giraud, E., Klaver, G., Negrel, P., 2009. Natural versus anthropogenic Pharmacology and toxicology of terbium, thulium, and ytterbium chlorides. sources in the surface- and groundwater dissolved load of the Dommel river Toxicol. Appl. Pharmacol. 5, 427–436. (Meuse basin): constraints by boron and strontium isotopes and gadolinium Haley, T.J., Komesu, N., Efros, M., Koste, L., Upham, H.C., 1964b. Pharmacology and anomaly. J. Hydrol. 369, 336–349. toxicology of praseodymium and neodymium chlorides. Toxicol. Appl. Phar- Rabiet, M., Brissaud, F., Seidel, J.-L., Pistre, S., Elbaz-Poulichet, F., 2005. Deciphering macol. 6, 614–620. the presence of wastewater in a medium-sized Mediterranean catchment Haley, T.J., Koste, L., Komesu, N., Efros, M., Upham, H.C., 1966. Pharmacology and toxicology of dysprosium, holmium, and erbium chlorides. Toxicol. Appl. using a multitracer approach. Appl. Geochem. 20, 1587–1596. Pharmacol. 8, 37–43. Rabiet, M., Brissaud, F., Seidel, J.L., Pistre, S., Elbaz-Poulichet, F., 2009. Positive Haley, T.J., Upham, H.C., Raymond, K., Komesu, N., 1961. Toxicological and gadolinium anomalies in wastewater treatment plant effluents and aquatic pharmacological effects of gadolinium and samarium chlorides. Br. J. Pharm- environment in the Herault watershed (South France). Chemosphere 75, col. Chemother. 17, 526–532. 1057–1064. Hammond, C.R., 2011. The elements. In: Haynes, W.M. (Ed.), CRC Handbook of Schlosser, C., Croot, P.L., 2008. Application of cross-flow filtration for determining Chemistry and Physics, 92nd ed. CRC Press, London, p. 2656. the solubility of iron species in open ocean seawater. Limnol. Oceanogr. Humphries, M., 2011. Rare Earth Elements: The Global Supply Chain. CRS Report Methods 6, 630–642. for Congress R41347. Congressional Research Service, Washington DC. Shi, W.L., Shen, X.Y., Ma, X.Y., 2006a. Effects of samarium on liver and kidney of /http://www.fas.org/sgp/crs/natsec/R41347.pdfS (accessed 31.10.12). rats. J. Rare Earths 24, 415–418. Knappe, A., Moller,¨ P., Dulski, P., Pekdeger, A., 2005. Positive gadolinium anomaly Shi, W.L., Shen, X.Y., Ma, X.Y., 2006b. Influence of samarium on learning and in surface water and ground water of the urban area Berlin, Germany. Chem. memory function of rats. J. Rare Earths 24, 419–422. Erde 65, 167–189. Sholkovitz, E.R., 1992. Chemical evolution of rare earth elements: fractionation Kulaksız, S., Bau, M., 2007. Contrasting behaviour of anthropogenic gadolinium between colloidal and solution phases of filtered river water. Earth Planet. Sci. and natural rare earth elements in estuaries and the gadolinium input into the Lett. 114, 77–84. North Sea. Earth Planet. Sci. Lett. 260, 361–371. Sholkovitz, E.R., 1995. The aquatic chemistry of rare earth elements in rivers and Kulaksız, S., Bau, M., 2011a. Rare earth elements in the Rhine River, Germany: first estuaries. Aquat. Geochem. 1, 1–34. case of anthropogenic lanthanum as a dissolved microcontaminant in the Sun, H., Wang, X., Hua, Z., Wu, C., Wang, L., Dai, L., Li, Z., Chen, Y., 1996. hydrosphere. Environ. Int. 37, 973–979. Bioconcentration and elimination of five light rare earth elements in carp Kulaksız, S., Bau, M., 2011b. Anthropogenic gadolinium as a microcontaminant in (Cyprinus carpio L.). Chemosphere 33, 1475–1483. tap water used as drinking water in urban areas and megacities. Appl. Tricca, A., Stille, P., Steinmann, M., Kiefel, B., Samuel, J., Eikenberg, J., 1999. Rare Geochem. 26, 1877–1885. earth elements and Sr and Nd isotopic compositions of dissolved and Kunnemeyer,¨ J., Terborg, L., Meermann, B., Brauckmann, C., Moller,¨ I., Scheffer, A., suspended loads from small river systems in the Vosges mountains (France), Karst, U., 2009. Speciation analysis of gadolinium chelates in hospital effluents the river Rhine and groundwater. Chem. Geol. 160, 139–158. and wastewater treatment plant sewage by a novel HILIC/ICP-MS method. Verheul, M., Klaver, G., Bakker, I., Petelet-Giraud, E, 2011. The Distribution of the Environ. Sci. Technol. 43, 2884–2890. Lanthanum Anomaly in the Dutch Branches of the Rhine River. Abstract H41H- Lawrence, M.G., 2010. Detection of anthropogenic gadolinium in the Brisbane 1151 Presented at 2011 Fall Meeting, 5–9 December, AGU, San Francisco, River plume in Moreton Bay, Queensland, Australia. Mar. Pollut. Bull. 60, California. 1113–1116. Verplanck, P.L., Taylor, H.E., Nordstrom, D.K., Barber, L.B., 2005. Aqueous stability Lawrence, M.G., Jupiter, S.D., Kamber, B.S., 2006. Aquatic geochemistry of the rare of gadolinium in surface waters receiving sewage treatment plant effluent, earth elements and yttrium in the Pioneer River catchment, Australia. Mar. Boulder Creek, Colorado. Environ. Sci. Technol. 39, 6923–6929. Freshwater Res. 57, 725–736. Weiss, G.B., Goodman, F.R., 1975. Interactions between several rare-earth ions and Lawrence, M.G., Kamber, B.S., 2007. Rare earth element concentrations in the Natural Water Reference Materials (NRCC) NASS-5, CASS-4 and SLEW-3. calcium-ion in vascular smooth-muscle. J. Pharmacol. Exp. Ther. 195, 557–564. Geostand. Geoanal. Res. 31, 95–103. Zhang, H., He, X., Bai, W., Guo, X., Zhang, Z., Chai, Z., Zhao, Y., 2010. Ecotoxicological McLennan, S.M., 1989. Rare earth elements in sedimentary rocks; influence of assessment of lanthanum with Caenorhabditis elegans in liquid medium. provenance and sedimentary processes. Rev. Mineral. Geochem. 21, 169–200. Metallomics 2, 806–810. Moller,¨ P., Dulski, P., Bau, M., Knappe, A., Pekdeger, A., Sommer-von Jarmersted, C., Zhu, Y., Hattori, R., Rahmi, D., Itoh, S.O., Fujimori, E., Umemura, T., Haraguchi, H., 2000. Anthropogenic gadolinium as a conservative tracer in hydrology. 2005. Fractional distributions of trace metals in surface water of Lake Biwa as J. Geochem. Explor. 69–70, 409–414. studied by ultrafiltration and ICP-MS. Bull. Chem. Soc. Jpn. 78, 1970–1976. Moller,¨ P., Morteani, G., Dulski, P., 2003. Anomalous gadolinium, cerium, and Zhu, Y., Hoshino, M., Yamada, H., Itoh, A., Haraguchi, H., 2004. Gadolinium yttrium contents in the Adige and Isarco river waters and in the water of their anomaly in the distributions of rare earth elements observed for coastal tributaries (Provinces Trento and Bolzano/Bozen, NE Italy). Acta Hydrochim. seawater and river waters around Nagoya City. Bull. Chem. Soc. Jpn. 77, Hydrobiol. 31, 225–239. 1835–1842.

137

138 CHAPTER VI – ANTHROPOGENIC GADOLINIUM AS A

MICROCONTAMINANT IN TAP WATER USED AS DRINKING WATER

IN URBAN AREAS AND MEGACITIES

This chapter was published in Vol. 26 of the journal Applied Geochemistry in 2011 and can be found online at http://dx.doi.org/10.1016/j.apgeochem.2011.06.011.

139 Applied Geochemistry 26 (2011) 1877–1885

Contents lists available at ScienceDirect

Applied Geochemistry

journal homepage: www.elsevier.com/locate/apgeochem

Anthropogenic gadolinium as a microcontaminant in tap water used as drinking water in urban areas and megacities ⇑ Serkan Kulaksız, Michael Bau

Earth and Space Sciences, School of Engineering and Science, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany article info abstract

Article history: Gadolinium chelates have been used since 1988 as contrast agents in magnetic resonance imaging (MRI), Available online 17 June 2011 and produce positive anthropogenic Gd anomalies in rare earth element (REE) patterns of river and lake waters. These Gd compounds are not removed in wastewater treatment plants (WWTP) due to their high stabilities, and are transferred to surface waters with the clearwater discharge from WWTP. Through nat- ural and induced bank filtration, the anthropogenic Gd is also transported into groundwater. To date, there are no related acute health risks known, but the potential long-term effects of exposure to low doses have not been studied. Here REE data is presented for tap water from the City of Berlin, Germany, a metropolitan area that is known for its anthropogenic Gd-rich rivers and groundwater. Natural and induced bank filtration play important roles in Berlin’s freshwater resource management. Therefore, the extent to which municipal tap water that is used as drinking water is affected by anthropogenic Gd was investigated. Large positive Gd anomalies were found in tap water samples from the western districts of Berlin, indicating the pres- ence of up to 18 ng/L of anthropogenic Gd on top of a geogenic background of 0.54 ng/L. In marked con- trast, the amount of anthropogenic Gd in tap water from the eastern districts of Berlin is negligible to minor (maximum of 0.18 ng/L on top of a geogenic background of 0.26 ng/L). This strong regional differ- ence likely results from the specific historical situation of Berlin, where before the re-unification of Ger- many in 1990, natural and induced bank filtration were necessities in isolated West Berlin, but unimportant in East Berlin, a situation that has seen little change during the past 20 years. Thus, drinking water resources in the western part of Berlin are more strongly affected by anthropogenic Gd than those in the eastern part. The high anthropogenic Gd concentrations found in some tap waters in Berlin clearly show that the Gd initially used as contrast agent is removed neither during natural nor artificial water treatment. This is further evidence for the high stability and long environmental half-lives of these com- pounds. Considering that the amount of anthropogenic Gd in the Havel River in Berlin has increased more than 4-fold over the past 15 years and that water migration from the Havel River to the groundwater wells take years to decades, the amounts of anthropogenic Gd in West Berlin tap water will increase fur- ther over the next few years. Due to its presence in tap water that is consumed as drinking water, millions of people are exposed to low doses of these anthropogenic Gd chelates. Additional data for the City of London, UK, for example, indicate that this is not a local phenomenon confined to the City of Berlin, but rather a common feature of tap water in metropolitan areas and megacities in countries with highly developed health care systems. Hence, the REE distribution in tap waters used for human consumption should be monitored, especially since the anthropogenic Gd chelates can also be used as tracers for emerging microcontaminants such as steroids, pharmaceuticals and personal care products. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction ior of individual elements easy to discern. Cerium and Eu anoma- lies result from the specific redox chemistry of these elements, The rare earth elements (REE) comprise the elements La and Ce while very small anomalies for La, Gd and Lu stem from slight dif- to Lu and behave coherently in natural systems. Their coherence is ferences between the stabilities of chemical REE complexes (e.g., commonly illustrated using plots for which measured REE concen- Bau, 1999). These anomalies are natural and not the result of trations are normalized to those of a natural reference material. anthropogenic activity. Such REE plots reveal smooth patterns, making anomalous behav- The occurrence of anomalously high concentrations of Gd in the environment was first reported for rivers in Germany in 1996 (Bau ⇑ Corresponding author. and Dulski, 1996). Since then, the presence of large positive Gd E-mail address: [email protected] (M. Bau). anomalies of varying size has also been reported for other rivers

0883-2927/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apgeochem.2011.06.011 140 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885 in Europe (Tricca et al., 1999; Elbaz-Poulichet et al., 2002; Möller Künnemeyer et al., 2009) and in urban WWTP effluent (Fig. 2; et al., 2002, 2003; Knappe et al., 2005; Rabiet et al., 2005, 2009; Bau and Dulski, 1996; Möller et al., 2002; Verplanck et al., 2005, Morteani et al., 2006; Kulaksız and Bau, 2007), Asia (Nozaki 2010; Bau et al., 2006; Kulaksız and Bau, 2007; Lawrence et al., et al., 2000; Zhu et al., 2004, 2005), North America (Verplanck 2009; Petelet-Giraud et al., 2009; Rabiet et al., 2009). Künnemeyer et al., 2005; Bau et al., 2006) and Australia (Lawrence et al., et al. (2009) have directly confirmed the presence of different Gd- 2006). Fig. 1 shows selected rivers with and without positive Gd CA through a speciation analysis in hospital and WWTP effluent. It anomalies. The largest positive Gd anomaly in rivers observed to is via the discharge of WWTP effluent that surface waters receive date has been found in the Havel River at and downstream from anthropogenic Gd and develop anthropogenic positive Gd anoma- Berlin, Germany. The Havel River is a low-discharge river receiving lies. Verplanck et al. (2010) conducted a detailed study on the a large amount of waste water treatment plant (WWTP) effluent behavior of Gd through large metropolitan WWTP in the USA, (Bau and Dulski, 1996). These Gd anomalies are of anthropogenic showing the persistence of Gd-CA in the dissolved phase rather origin, as it is impossible to produce individual REE fractionations than transferring into the sludge. This behavior is not observed of this extent by natural processes in pristine environments (Figs. 1 in WWTP that utilize reverse osmosis membranes, where 99.85% and 2). of anthropogenic Gd is removed in the so-called Advanced Water The source of anthropogenic Gd is the widespread application of Treatment Plants of SE Queensland, Australia (Lawrence et al., Gd-chelates as contrast agents in clinical and diagnostic medical 2010). imaging. Since the approval of the first Gd-based contrast agent Previously, the authors reported on the presence and behavior (Gd-DTPA) for clinical use in 1988 (both in Germany and the of anthropogenic Gd in rivers in NW Germany, in the Weser River USA), the number of applications has increased significantly. By estuary (Germany) and the southern North Sea (Kulaksız and Bau, September 2009, an estimated 100 million applications of Gd- 2007). The anthropogenic Gd compounds behave conservatively in DTPA had been performed worldwide. In 2005 alone, more than estuaries as river water is admixed with seawater. This is in 20 million Gd-based contrast enhancement procedures were per- marked contrast to the behavior of the geogenic (natural) Gd formed worldwide (Idée et al., 2006). An estimated 22–66 tons of and that of the other REE, which because of their high particle- Gd is used each year worldwide for contrast agent applications reactivity are removed from solution together with colloids that (20 million applications per year, each using 1.1–3.3 g of Gd). This aggregate due to salting out and are deposited in the estuary. Thus, is a sizeable fraction of the annual total world production of Gd the Gd anomaly is carried through estuaries into coastal waters (400 tons). The use of Gd for this purpose relies on its paramag- and semi-closed sea-basins. The conservative behavior of Gd-CA netic properties. However, free Gd3+ is highly toxic and is known to was also confirmed for the Brisbane River plume in Queensland, affect Ca chemistry and disrupt cellular processes, for example Australia (Lawrence, 2010). These findings suggest that anthropo- (Yang and Sachs, 1989; Biagi and Enyeart, 1990; Molgo et al., genic Gd occurs as highly stable unreactive compounds that have 1991; Shellock and Spinazzi, 2008). Hence, chelation of Gd in the long environmental half-lives. Furthermore Möller and Dulski MRI contrast media is a prerequisite for its safe use, and enhances (2010a,b) have studied transmetallation of Gd-DTPA (Gd-diethyl- the rapid removal of the contrast agent from the body after admin- enetriaminepentaacetic acid or gadopentetic acid) with other istration. The chelated forms of Gd (Gd contrast agents – Gd-CA) REE, as well as Cu and Y, in suspension experiments with clay min- are used in relatively high doses of 1.1–3.3 g Gd per average adult erals. Among the findings was an increase of REE-DTPA in solution patient (0.1–0.3 mmol per kg body weight, Bayer HealthCare AG, and the removal of Gd from solution due to absorption of free Gd 2009). In patients with functional kidneys the Gd-CA are rapidly after some of the Gd in Gd-DTPA had been replaced by other removed from the body through urine. Hence, Gd-CA have been REE. They estimated that about 10% of the Gd-DTPA undergoes found in hospital effluent (Kümmerer and Helmers, 2000; transmetallation into REE-DTPA per year, suggesting that

10 Susquehanna R. ('02) USA Lake Erie ('02) USA 100 Ohio R. ('02) USA R. Thames ('09) GBR Danube R. ('05) AUT Rhine R. ('08) GER Havel R. ('09) GER 6 10 1 Havel R. ('95) GER x 10

/ PAAS 1 E

RE 0.1 Västerdalälven R. ('94) SWE Toshibetsu R. ('95) JPN Spring Cr. ('02) USA 0.1 Delaware R. ('02) USA A Wiembach Cr. ('04) GER B

0.01 Er Pr Pr Er Lu La La Lu Tb Tb Dy Dy Yb Eu Yb Eu Ho Ce Nd Ce Nd Ho Gd Gd Tm Tm Pm Sm Pm Sm

Fig. 1. REESN patterns of water samples from selected rivers (R.) and creeks (Cr.) without (A) and with (B) anthropogenic Gd input. Data for the rivers Västerdalälven (Sweden), Toshibetsu (Japan) and Havel (’95 – Germany) from Bau and Dulski (1996); data for Spring Creek and rivers Delaware, Susquehanna, Ohio and Lake Erie (USA) from Bau et al. (2006); all other data from this study. 141 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885

of anthropogenic sources. Morteani et al. (2006) have also used Denver, Colorado ('07) USA anthropogenic Gd as a tracer of sewage water in surface water. Berlin ('96) GER Bremen ('06) GER Despite the widely reported occurrence of anthropogenic Gd in 100 Vienna ('07) AUT surface water and groundwater, evidence of anthropogenic Gd in Prague ('02) CZE tap water is still scarce. Published REE data are mostly for water Mackay ('05) AUS from production wells, aquifers and treated groundwater (deBoer et al., 1996; Janssen and Verweij, 2003; Rabiet et al., 2005, 2006; 6

10 Biddau et al., 2009). Since the first reported case of Gd-CA in a sin- 10 x gle tap water sample (Bau and Dulski, 1996), only a few studies

AS have reported Gd-CA in municipal tap water (Möller et al., 2002) and in wells used for drinking water production (Möller et al., 2000, 2002; Rabiet et al., 2005, 2006). The lack of data along with

E/PA 1 increasing concern about the potential toxicity related to Gd-CA

RE applications (to be addressed at the end of this contribution) has prompted a study of tap water from several districts in the City of Berlin, Germany. Aside from the already known adverse health 0.1 effects of high doses of Gd-CA, there may be unknown long-term effects of low-dosage exposure to Gd-CA, especially for infants and pregnant women. Hence, the extent to which anthropogenic Gd in surface and groundwater can be traced in municipal tap water that is distributed as drinking water was investigated. Pr Er La Lu Tb Dy Eu Yb Ce Nd Ho Gd Tm Pm Sm

Fig. 2. REESN patterns of WWTP effluents showing extremely large positive Gd 2. Materials and methods anomalies due to anthropogenic Gd input. Data for Denver, Colorado, USA, from Verplanck et al. (2010); Berlin, Germany, from Bau and Dulski (1996); Prague, Czech Republic, from Möller et al. (2002), and Mackay, Australia, from Lawrence et al. 2.1. Sampling (2006; Eu interpolated using Sm and Tb); Bremen, Germany, from Kulaksiz and Bau (2007); Vienna, Austria, from this study. Sampling locations, pH and conductivity are listed in Table 1. Twenty-three tap water samples from different districts of the City equilibrium will only be reached after 70 years (Möller and Dulski, of Berlin, together with a shallow ground water sample from the 2010a,b). eastern part of Berlin were investigated. The Havel River receives Anthropogenic Gd has also been reported in groundwater (Möl- high anthropogenic Gd input from WWTP effluent and was also ler et al., 2000; Knappe et al., 2005; Rabiet et al., 2009), indicating sampled for comparison with published data (Bau and Dulski, that processes such as natural or induced bank filtration do not 1996) from the mid-1990s. prevent the migration of anthropogenic Gd into aquifers (Fig. 3). All samples were collected in acid-cleaned 1 L polyethylene bot- Knappe et al. (2005) applied anthropogenic Gd present in ground- tles after rinsing with sample volume several times. Water from water as a tracer for surface water influence, while Petelet-Giraud each tap ran for at least 10 min before samples were taken. Sam- et al. (2009) have successfully combined REE data with Sr and B ples were filtered (<0.2 lm) and acidified to pH 1.8–2.0 using isotopes to discriminate between natural sources and two types ultrapure HCl. In order to be able to reliably quantify all REE in tap water in spite of their low concentrations, samples were pre- concentrated before measuring with inductively coupled plasma mass spectrometry (ICPMS). Thulium (Tm) was used for checking recovery rates of REE. All samples have recovery rates higher than 10 94%. The pre-concentration method followed the procedure of Bau and Dulski (1996) who adjusted the approach of Shabani et al. (1992) to optimize the pre-concentration of REE from freshwater.

6 2.2. Quantification of anomalies 10 1 Rare earth element anomalies are calculated based on the coherent behavior of the REE. Smooth REE patterns are produced for pristine (anthropogenically unaffected) samples by normalizing measured concentrations to those of a known reference material EE / PAAS x 0.1

R (e.g. shale). This approach facilitates the approximation of the background concentration for any anomalous REE via interpola- tion. REE patterns presented in this contribution are normalized West Berlin Shallow Groundwater1 to Post-Archean Average Australian Shale (PAAS) from McLennan West Berlin Deep Groundwater1 (1989). The shale-normalized measured element concentration 0.01 East Berlin Shallow Groundwater (Wuhlheide) (REESN) is then divided by this calculated normalized background concentration (REESN) to give the REESN/REESN ratio. There are sev- eral methods to estimate the background concentration of a given Pr Er La Lu Tb Dy Eu Yb Ce Nd Ho Gd Tm Pm Sm element, depending on the composition of the individual sample. In river waters and ground waters, Gd behaves as a light REE (LREE) Fig. 3. REE patterns for groundwater wells in the western part of Berlin, Germany, SN (Bau et al., 2006) and background Gd levels in such waters are showing an anthropogenic Gd anomaly (1data from Knappe et al. (2005)) and very shallow groundwater from Wuhlheide (this study) in the eastern part of Berlin, that therefore best estimated via linear regression using all suitable is devoid of anthropogenic Gd input. LREE. Note that this approach differs significantly from that com- 142 Table 1 ICP-MS data for tapwater, groundwater and surface water from Berlin. All concentrations in pmol/L unless otherwise noted.

⁄ ⁄ Sample Name Post Code La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu RREE Rb Sr Ba U pH conductivity GdSN =GdSN Gd Gdanthr %Gdanthr (nmol/L) (lmol/L) (lmol/L) (nmol/L) (lS/cm) Mitte 10117 8.42 9.60 2.37 12.8 3.65 0.87 44.7 0.90 6.66 1.56 5.91 1.07 9.31 2.07 110 27 2.8 0.45 1.69 7.51 784 8.00 5.59 39 87 Friedrichshain 10247 3.76 5.38 1.32 6.78 1.59 0.46 3.03 0.54 4.55 1.24 5.45 1.16 11.9 2.93 50.1 19 3.7 0.78 1.10 7.50 954 1.46 2.07 0.96 32 Reichstag 10557 3.88 7.94 1.41 6.48 2.23 0.65 115 0.70 6.24 1.65 6.81 1.25 11.1 2.35 168 39 3.4 0.45 1.17 7.46 - 33.7 3.42 112 97

Reichstag 2 10557 4.11 7.29 1.43 6.37 2.01 0.62 60.4 0.66 5.36 1.51 5.97 1.18 11.2 2.45 111 31 2.1 0.41 2.23 7.50 776 21.2 2.84 58 95 1877–1885 (2011) 26 Geochemistry Applied / Bau M. Kulaksız, S. Jungfernheide 10589 3.96 7.43 1.10 5.04 1.26 0.35 3.50 0.33 3.17 0.92 3.55 0.61 5.07 1.02 37.3 20 2.0 0.29 0.82 7.51 626 2.22 1.58 1.92 55 Zoologischer Garten 10623 5.48 8.60 1.69 8.44 2.99 0.84 74.3 0.84 6.56 1.76 6.90 1.22 10.4 2.11 132 29 2.1 0.40 2.10 7.43 766 15.0 4.96 69 93 Kurfürstendamm 10789 3.43 5.03 0.98 5.18 2.05 0.61 38.0 0.69 4.98 1.33 5.04 0.94 8.42 1.82 78.5 26 2.0 0.36 1.72 7.52 716 9.98 3.81 34 90 Schöneberg 10829 4.33 5.94 1.14 6.30 1.95 0.57 16.9 0.53 4.52 1.22 4.59 0.82 7.05 1.38 57.2 22 2.0 0.32 1.14 7.60 706 5.26 3.20 14 81 Neukölln 12055 3.94 4.84 1.08 5.72 2.07 0.58 30.6 0.62 4.96 1.40 4.70 0.88 7.89 1.70 71.0 24 2.0 0.35 1.39 7.45 717 8.39 3.65 27 88 Alt-Mariendorf 12107 2.38 4.40 0.68 3.06 0.92 0.25 11.3 0.23 1.82 0.51 1.90 0.32 2.56 0.54 30.9 16 4.0 0.58 0.24 7.52 - 8.92 1.27 10.0 89 Steglitz 12165 3.66 4.08 0.60 2.92 1.00 0.30 7.44 0.30 1.72 0.47 1.70 0.28 2.21 0.47 27.1 13 2.1 0.45 0.27 7.55 733 4.63 1.61 5.83 78 Adlershof 12489 3.77 5.14 0.99 5.53 1.51 0.43 3.50 0.50 3.18 0.89 2.91 0.46 3.45 0.80 33.1 29 7.2 0.56 0.36 7.45 - 1.52 2.30 1.20 34 Friedrichshagen 12587 5.76 7.66 1.46 5.83 1.36 0.34 2.26 0.36 2.80 0.76 2.96 0.57 5.21 1.11 38.4 20 3.5 0.52 0.84 7.49 404 1.54 1.47 0.80 35 Kaulsdorf 12621 2.48 3.24 0.75 3.77 1.10 0.33 2.76 0.40 2.41 0.65 2.29 0.44 4.03 0.92 25.6 23 4.0 0.54 0.88 7.30 816 1.70 1.63 1.13 41 Marzahn 12679 2.38 4.33 0.97 5.26 1.63 0.41 3.20 0.47 3.70 0.91 3.05 0.53 4.51 0.99 32.3 23 4.0 0.59 0.94 7.54 900 1.21 2.64 0.56 18 143 Buch 13125 3.59 7.98 1.14 5.52 1.75 0.40 3.00 0.48 3.73 0.92 3.42 0.61 5.27 1.06 38.9 21 3.4 0.46 0.64 7.53 790 1.13 2.65 0.35 12 Pankow 13187 6.49 11.6 2.37 11.3 3.20 0.77 4.23 0.75 5.58 1.36 4.81 0.80 6.42 1.21 60.9 17 2.4 0.37 0.58 7.43 680 0.95 4.45 -0.22 - Wedding 13347 5.24 8.64 1.44 7.07 2.38 0.65 51.6 0.76 5.92 1.63 6.54 1.22 11.1 2.40 107 33 2.2 0.44 1.95 7.58 760 13.6 3.78 48 93 Wittenau 13437 6.11 8.01 1.84 9.06 3.00 0.71 53.8 0.87 6.72 1.81 6.52 1.23 11.3 2.40 113 34 2.3 0.46 2.07 7.44 776 11.4 4.74 49 91 Tegel 13507 4.09 7.12 1.15 7.00 2.35 0.73 58.7 0.83 5.97 1.56 6.11 1.17 10.7 2.40 110 32 2.2 0.43 1.76 7.47 772 13.5 4.35 54 93 Spandau 13597 6.76 8.76 1.54 7.69 1.97 0.53 3.58 0.56 4.16 1.03 3.84 0.64 5.23 1.06 47.4 19 2.1 0.35 0.62 7.51 724 1.35 2.65 0.93 26 Zehlendorf 14169 1.62 2.82 0.48 2.27 0.86 0.21 10.1 0.24 1.70 0.47 1.51 0.28 2.44 0.49 25.5 12 2.1 0.48 0.21 7.45 731 7.05 1.44 8.70 86 Hohenzollerndamm 14199 4.03 6.06 1.42 5.94 1.43 0.49 9.2 0.49 3.52 0.95 3.56 0.58 4.6 0.98 43.3 21 4.0 0.53 0.26 7.44 – 5.68 1.63 7.62 82 Wuhlheide Groundwater 12459 17.5 34.6 4.80 22.9 5.81 1.49 8.42 1.87 20.6 8.51 45.6 11.0 129 32.1 344 36 2.8 0.16 12.2 7.42 1980 1.12 7.54 0.88 10 London Tapwater 11.8 20.8 4.30 20.6 5.81 1.32 12.3 1.20 8.99 2.48 9.62 1.80 16.3 3.16 121 24 2.7 0.13 2.41 – – 1.52 8.10 4.2 34 River Thames 24.4 41.8 8.25 39.0 9.20 2.36 28.0 1.67 12.4 3.37 14.5 2.78 25.7 4.96 218 34 2.8 0.13 2.18 – – 2.47 11 17 59 Danube River 65.2 88.4 17.3 78.9 18.1 4.43 48.6 3.22 21.0 4.48 14.2 2.04 14.1 2.23 406 20 2.5 0.19 4.22 – – 2.29 21 27 56 Havel River 28.8 15.0 4.69 20.1 4.43 0.93 3136 0.87 6.10 1.51 5.22 0.81 6.05 1.20 3231 118 3.1 0.27 1.50 – – 644 4.87 3131 100 Havel River (’95) A 15.8 24.0 4.08 17.3 4.56 0.91 675 1.02 7.43 1.96 8.36 1.49 13.8 - – – – – – – – 121 5.58 669 99

Tm⁄ = Tm concentrations interpolated using Er and Yb. The only exception is the Havel River (‘95) where measured Tm concentration is shown. RREE = Sum of REE including Tm ⁄. GdSN /GdSN = Gadolinium anomaly calculated according to Eqs. (1) and (2) . Gdanthr (pM) = Anthropogenic Gd concentration according to Eqs. (1)–(3) . %Gdanthr = Anthropogenic Gd content as a percentage of total Gd. A Bau and Dulski 1996 . S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885 monly used in REE geochemistry. The elements Pr, Nd and Sm do 3.1. Distribution of uranium, barium, rubidium and strontium not carry anomalies and do not differ significantly from the general linear trend of the LREE. Lanthanum, Ce and Eu are potentially The median U concentration in the tap water samples is anomalous with respect to the other LREE and are not included 0.94 nmol/L (0.22 lg/L), while individual U concentrations range in the linear analysis for calculating background (geogenic) Gd from 0.21 nmol/L (0.05 lg/L) to 2.23 nmol/L (0.53 lg/L). The high- concentrations. Eqs. (1a)–(1d) clarify the least squares method em- est U concentration was measured in the shallow groundwater ployed for linear regression: sample at Wuhlheide (12.2 nmol/L, 2.90 lg/L). Uranium in tap water has recently become a concern in Germany, as tap water ¼ þ ð Þ log GdSN b0 b1xGd 1a in some areas shows U concentrations as high as 38.95 lg/L or 30.12 lg/L (towns of Maroldsweisach and Fürfeld, respectively), where b0 and b1 are coefficients defined by: P i.e. well above recommended values (Schnug et al., 2008). Uranium ½ðx xÞðy yÞ is a toxic metal that may pose health problems unrelated to radio- b ¼ Pi i ð1bÞ 1 2 activity. Uranium concentrations in the samples presented here are ½ðxi xÞ much lower than guideline values set to date such as the 63 nmol/L and (15 lg/L) suggested by the World Health Organization (2008). However, a lower limit of 10 lg/L has also been suggested by the ð1cÞ ̅ German Drinking Water Commission (TWK) of the Federal

Overbars for x and y signify average values. In calculating b1, the Ministry of Health, located at the Federal Environment Agency x value for each element must be picked so that the differences in x (UBA) (2008). An even lower limit of 2 lg/L is suggested for water values reflect the relative distances of the points along the x-axis that is used for preparing baby food by The Federal Institute of Risk on a REE plot (e.g. xPr =3, xNd =4, xSm = 6 and xGd = 8). Expanding Assessment (2006). Samples from the western districts of Berlin Eq. (1b), we get: show higher U concentrations (median 1.17 nmol/L or 0.29 lg/L)

½ð Þð Þþð Þð Þþð Þð Þ ¼ xPr x log PrSN y xNd x log NdSN y xSm x log SmSN y ð Þ b1 2 2 2 1d ðxPr xÞ þðxNd xÞ þðxSm xÞ

x þ x þ x log Pr þ log Nd þ log Sm than those from the eastern districts (0.86 nmol/L or 0.20 lg/L), where x ¼ Pr Nd Sm and y ¼ SN SN SN. 3 3 although the two lowest samples are from the SW of Berlin (e.g. Once the coefficients b0 and b1 are known, Eq. (1a) yields the Zehlendorf (0.21 nmol/L or 0.05 lg/L). The median values are simi- logarithmic shale normalized background Gd concentration lar to data from Schnug et al. (2008) for Berlin and to the median ð Þ logGdSN . Raising 10 to this value yields the shale normalized value for Germany of 1.34 nmol/L or 0.32 lg/L (Birke et al., 2009). background Gd concentration GdSN, extrapolated using the ele- Median dissolved concentrations of Ba, Rb and Sr are 0.45 lmol/L ments Pr, Nd and Sm. (0.06 mg/L), 23.5 nmol/L (2.01 mg/L) and 2.31 lmol/L (0.20 mg/L), The anomaly is then quantified by the ratio GdSN/GdSN, where respectively. These values are well below any drinking water limits, GdSN is the shale normalized total measured Gd concentration. e.g. Ba: 5.1 lmol/L (0.7 mg/L) (World Health Organization, 2008). Using GdSN calculated using Eq. (1a), the geogenic background con- centration Gd⁄ is calculated using Eqs. (1) and (2): 3.2. Rare earth elements ¼ ½ ð Þ Gd GdSN Gd SN 2 The total dissolved REE content in our samples has a median of where [Gd]SN denotes the Gd concentration of the PAAS reference 50.1 pmol/L (7.94 ng/L) and ranges from 26 pmol/L (4.0 ng/L) to shale (4.66 mg/kg). Extrapolated Gd⁄ is subtracted from total Gd measured to yield the anthropogenic Gd concentration (Gdanthr): Buch Gdanthr ¼ Gdtotal Gd ð3Þ Pankow 10 Friedrichshagen In this study, a positive anthropogenic Gd anomaly is defined by Marzahn aGdSN/GdSN ratio above unity, since Gd behaves as a LREE and no Kaulsdorf HREE are used in estimating background Gd. Other authors have Treptower Park used Sm and Tb for interpolating background Gd concentrations, 6 Adlershof 0 requiring cutoff GdSN/GdSN ratios larger than unity, in order to offset 1 x1 the increasing effect of Tb (a HREE) on background Gd concentra- tions. The method utilizing Pr, Nd and Sm for extrapolation elimi- AAS nates this problem, although the extrapolation can amplify small P

changes due to the relative distance of the elements with respect E/ to the element of interest (Gd) along a theoretical straight line. 0.1 RE

3. Results and discussion

Dissolved concentrations of REE, Ba, Rb, Sr and U are shown in 0.01 Table 1. The sample set shows a narrow range of pH values from 7.3 to 7.6 and conductivity values between 404 and 954 lS/cm (Ta- ble 1). There are no systematic differences between the samples in Pr Er Lu La Tb Dy Yb Eu Ce Nd Ho Gd Tm terms of pH and conductivity. The small variability in pH values al- Pm Sm lows straightforward comparison of dissolved REE levels from all Fig. 4. REESN patterns of tap water samples from eastern districts of Berlin, sampling sites. Germany, lacking anthropogenic positive Gd anomalies. 144 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885

Hohenzollerndamm Tegel Zoologischer Garten Alt-Marienhof 10 Jungfernheide Wittenau 10 Wedding Kurfürstendamm Neukölln Schöneberg Zehlendorf Steglitz Reichstag

6 Spandau

0 Mitte Reichstag 2 1 1 London ('09) GBR AAS x 1 P

0.1 0.1 REE /

0.01 0.01 Pr Pr Er Er La La Lu Lu Tb Tb Dy Dy Yb Yb Eu Eu Ce Nd Ce Nd Ho Ho Gd Gd Tm Tm Pm Sm Pm Sm

Fig. 5. REESN patterns of tap water samples from western districts of Berlin, Germany and from London, UK. All samples show significant positive anthropogenic Gd anomalies with the exception of tap water from Spandau. Sample Mitte (although located east of the former Berlin Wall) is also included due to the large Gd anomaly it carries (for further explanations see text).

168 pmol/L (26.3 ng/L). These values are in close agreement with (17.6 ng/L), respectively, on top of a geogenic background of the limited amount of data published for Berlin tap water (Bau 1.61 pmol/L (0.25 ng/L) and 3.42 pmol/L (0.54 ng/L), respectively. and Dulski, 1996). In spite of the large variability of REE concentra- The only exceptions are samples from Spandau and Jungfernheide, tions in groundwater, it appears that dissolved REE concentrations which show relatively small anomalies (GdSN/GdSN: 1.35 and in Berlin tap water are low when compared to the limited data 2.22, respectively). available worldwide (e.g. Möller et al., 2002; Janssen and Verweij, The regional difference between the two subsets of samples is 2003; Lawrence et al., 2009). strikingly clear, both qualitatively (Figs. 4 and 5) and quantitatively The REE distribution in the present sample set is strongly LREE- (Table 1). The samples from the eastern districts of Berlin show a depleted with a median LaSN/LuSN ratio of 0.03. This is probably due median Gd anomaly of 1.49 while those from the western districts to the oxidative removal of Fe during water treatment, since pre- yield a median of 8.92, the latter indicating that measured Gd con- cipitation of Fe (oxy)hydroxides is known to preferentially remove centrations are on average one order of magnitude above geogenic LREE (e.g. Bau, 1999). The shallow groundwater sample from levels. Fig. 6 shows a map of Berlin with the locations of sampling

Wuhlheide is strongly enriched in heavy REE (HREE) (LaSN/ sites as vertical bars that are proportional to the size of the Gd LuSN = 0.005) and, in marked contrast to groundwater from the anomaly. western part of Berlin, does not carry anthropogenic Gd (Fig. 3; This strong difference between samples from the western and GdSN/GdSN = 1.12). Massmann et al. (2008) have studied the eastern districts of Berlin stems from underlying historical factors. groundwater ages in parts of Berlin using a multi-tracer approach. Before the reunification of Germany in 1990, West Berlin was an The bank filtrate in the vicinity of Lake Wannsee (one of the Havel enclave and groundwater supplies had to be managed very care- lakes) in western Berlin shows strong vertical stratification with fully with no supply available from surrounding areas. East Berlin respect to groundwater age that increases with increasing aquifer did not have such restrictions with respect to its drinking water re- depth and with increasing distance from the lake. However, based serves. This has led to the implementation of different water man- on 3H/3He dating, the average age of the groundwater at the pro- agement schemes on either side of the Berlin Wall. In contrast to duction wells is older than 10 years (Massmann et al., 2008). East Berlin, West Berlin used induced bank filtration in order to Fig. 4 shows REE patterns of tap water samples from the eastern keep groundwater levels constant. Today, about 70% of all drinking districts of Berlin. These samples show little to no anthropogenic water in Berlin relies on natural and induced bank filtration Gd with GdSN/GdSN ranging between 0.95 (Pankow) and 1.70 (Berliner Wasserbetriebe, 2008). The sites of groundwater recharge (Kaulsdorf). Quantification of the geogenic Gd background and in Berlin are shown in Fig. 6 (open squares) as is the former posi- the anthropogenic Gd concentrations from Eqs. (1)–(3) gives values tion of the Berlin Wall that separated West Berlin from the German that range from 1.47 pmol/L (or 0.23 ng/L) to 4.45 pmol/kg (or Democratic Republic (GDR). It is remarkable that even two decades 0.70 ng/L) of geogenic Gd and 0.35 pmol/kg (or 0.06 ng/L) to after reunification such a large difference exists in the anthropo- 1.20 pmol/L (or 0.19 ng/L) of anthropogenic Gd, respectively. Tap genic Gd content of tap water between the two parts of Berlin. water from Mitte is located in former East Berlin, but is not in- Determining the actual origin of the tap water at a given sample cluded in Fig. 4 for reasons that will be discussed later. location is not trivial, as water from several water works is mixed In western Berlin, the situation is completely different. In con- and detailed information is not readily available. Nevertheless, one trast to the samples shown in Fig. 4, tap water from Berlin’s wes- may speculate about the origin of the tap water based on the size tern districts shows large positive Gd anomalies (Fig. 5). These of the Gd anomaly and the REE pattern of a given sample, at least samples display Gd anomalies ranging from 4.63 (Steglitz)toas for the western districts. There are three major water works (WW) high as 33.7 (ironically a sample from the Reichstag, which houses operating in this part of Berlin: Tegel WW, Spandau WW and the Parliament of the Federal Republic of Germany) with anthropo- Beelitzhof WW (Fig. 6). The fact that the samples closest to the genic Gd concentrations of 5.83 pmol/L (0.92 ng/L) and 112 pmol/L Spandau WW (Spandau and Jungfernheide) have the smallest Gd 145 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885

Fig. 6. Map of the City of Berlin, Germany, with the tap water sampling sites indicated by bars proportional to the size of the Gd anomaly in the respective sample. The locations of water works are shown as light-gray circles with sizes proportional to the respective production volume. Light-gray squares indicate zones where surface water is used for induced bank filtration. The dashed line represents the former location of the Berlin Wall. Note that the presence of anthropogenic Gd in tap water is confined to the western districts of Berlin. Sample location for Reichstag is just west of the former border. anomalies (1.35 and 2.22, respectively) suggests that the water re- Wasserbetriebe, 2009). This causes a large strain on the drinking source(s) for Spandau WW which supplies 14% of Berlin’s tap water supplies of the city, and hence, bank filtration will in time water, is not as severely affected by anthropogenic Gd from surface play an even more important role in the water management water as the other WW. Tegel WW on the other hand produces 23% scheme of the city. Considering that the total number of Gd-CA of Berlin’s water supply and is the closest WW to the sample loca- applications will probably increase in the future as it did in the past, tions with the largest Gd anomalies. Sample locations closest to the the anthropogenic Gd content of surface and tap water can be ex- city centre all have large Gd anomalies and probably receive a sig- pected to increase further rather than decrease. This is already evi- nificant portion of their tap water from Tegel WW. Samples from dent in the GdSN/GdSN of the Havel River, which increased from 121 Zehlendorf, Steglitz and Alt-Mariendorf have similar total REE con- in 1995 (Bau and Dulski, 1996) to 644 in 2009 (Table 1). While the centrations and REE patterns and also the lowest U concentrations geogenic background in the Havel River remained at about the in the sample set. It is very likely that these samples receive their same level (5.58 pmol/L and 4.87 pmol/L), the amount of anthropo- tap water from the Beelitzhof WW. The fact that the very central genic Gd had increased more than 4-fold from 669 pmol/L in 1995 part of Berlin is characterized by large positive Gd anomalies, to 3131 pmol/L in 2009. Considering that it takes the Havel River regardless of whether the actual sampling site was located in for- water more than 10 years to reach the groundwater production mer East Berlin (Mitte) or West Berlin (Reichstag) suggests that dur- wells, this suggests that the anthropogenic Gd concentration of ing development of this area in the 1990s, these districts were Berlin’s tap water will increase further in the future. connected to the West Berlin water supply system. Due to the large Gd anomaly it carries, the sample Mitte has also been excluded 3.3. Toxicity of Gd3+ and Gd-CA from the East Berlin subset in terms of median and maximum Gd anomalies. The ionic radius of Gd3+ is very close to that of Ca2+, making The City of Berlin and its surrounding area has close to 4 million Gd3+ a strong inhibitor of Ca2+ activated enzymes as well as those inhabitants, each using on average 125 L of water per day (Berliner physiological processes that depend on Ca2+ to function (e.g., con- 146 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885 traction of smooth, skeletal and cardiac muscle, transmission of Reichstag (the location of the German parliament) and amounts nerve impulses; blood coagulation) (Evans, 1990). Gadolinium is to 112 pmol/L (17.6 ng/L) on top of a geogenic background concen- also an inhibitor of the reticuloendothelial system, while GdCl3 tration of 3.42 pmol/L (0.54 ng/L). The anthropogenic Gd in accumulates in Kupffer cells and leads to their death by inhibiting tap water from the western districts reflects its presence in phagocytic capacity (Evans, 1990). groundwater that receives anthropogenic Gd from surface water There is also a growing body of evidence for the adverse effects through bank filtration. Because of the specific historical situation of Gd-based contrast agents (Gd-CA) in the human body. The most of Berlin, (induced) bank filtration has always been important in eminent disorder is nephrogenic systemic fibrosis (NSF), a disease the water resource management of West Berlin, but did not play associated with the administration of Gd-CA in patients with renal an important role in East Berlin. Nevertheless, the presence of insufficiency (Broome, 2008, and references therein). First recog- anthropogenic Gd in tap water is by no means a local phenomenon. nized in 1997 and reported in 2000 (Cowper et al., 2000) NSF is The City of London, UK, faces similar challenges to Berlin (Table 1), the fibrosis of the skin and systemic tissues. By 2007, there were al- since the River Thames (GdSN/GdSN = 2.47) and tap water in London ready more than 400 cases of NSF worldwide that were directly (GdSN/GdSN = 1.52) also carry anthropogenic Gd (Figs. 1 and 5). It connected to exposure to Gd-CA (Thakral et al., 2007). Currently, can be expected that microcontaminants present in surface waters, there is no effective treatment for this potentially fatal illness, but such as contrast agents and pharmaceuticals, will eventually trans- there is research that aims at removing Gd-CA via dialysis and fer into the drinking water distributed in urban areas that derive hemoperfusion (Yantasee et al., 2010). As a result of the recent their freshwater from unconfined aquifers and rely on induced cases reported and research linking NSF to Gd-CA, all Gd-based con- bank filtration for groundwater recharge. This will be most pro- trast agent manufacturers in the USA have been required since 2007 nounced in megacities with highly evolved health care systems to include a box warning regarding the potential of the contrast where freshwater supplies are limited and per capita consumption agent for causing NSF. Idee et al. (2009) have summarized the avail- of pharmaceuticals and contrast agents is high. Tap water in such able information on the role of Gd-Ca in NSF development. There cities should, therefore, be monitored for the presence and behav- are several other adverse health effects of Gd-CA: transmetallation ior of anthropogenic Gd, especially since this is a rather fast and reactions may lead to the replacement of Gd3+ in the chelating cost-effective way to assess the potential presence of pharmaceu- agent by metals found in blood or extracellular fluids (Idée et al., ticals in drinking water. 2006, and references therein), although the results of Möller and Dulski (2010a,b) suggest that this might be a minor problem. However, there is also removal of Gd3+ from the chelating agent Acknowledgements during transport across the cellular membrane, concentrating Gd3+ inside cells (Cabella et al., 2006). Anomalously high concentra- We appreciate the help of B. Alexander, J. Mawick and D. Meissner tions of Gd have also been found in bone tissues, and are reported to (all Jacobs University Bremen). This work was partially sup- persist even 8 years after exposure (Darrah et al., 2009). Abraham ported by the German Science Foundation (DFG) through Grant et al. (2008) have found Gd in skin lesions of patients with NSF, No. BA 2289/2-1 to M.B. We also thank Jacobs University Bremen and demonstrated the in vivo release (through transmetallation) for the initial financial support of S.K. through its Integrated of free Gd3+ and its retention in different parts of the body. Ph.D. program. This contribution benefited from the reviews of However, at the concentrations found in Berlin tap water, Gd P.L. Verplanck and P. Möller and comments guest editor L.A. Munk. toxicity should not be a problem, considering that a person would have to drink 100 million L of tap water to match exposure levels References reached during a single MRI application. With the caveat that stud- ies on the effects of long-term exposure to low concentrations Abraham, J.L., Thakral, C., Skov, L., Rossen, K., Marckmann, P., 2008. Dermal especially on babies, infants and pregnant women do not exist, it inorganic gadolinium concentrations: evidence for in vivo transmetallation and seems unlikely that the anthropogenic Gd levels found in Berlin long-term persistence in nephrogenic systemic fibrosis. Brit. J. Dermatol. 158, 273–280. tap water impose a health risk on the more than 2 million people Bau, M., 1999. Scavenging of dissolved yttrium and rare earths by precipitating iron living in the western districts of the city. However, numerous phar- oxyhydroxide: experimental evidence for Ce oxidation, Y–Ho fractionation, and maceuticals can be expected to show similar environmental lanthanide tetrad effect. Geochim. Cosmochim. Acta 63, 67–77. Bau, M., Dulski, P., 1996. Anthropogenic origin of positive gadolinium anomalies in behavior as Gd-CA due to their high stability, water solubility river waters. Earth Planet. Sci. Lett. 143, 245–255. and ability to pass through WWTP unaffected or with little atten- Bau, M., Knappe, A., Dulski, P., 2006. Anthropogenic gadolinium as a micropollutant uation (Morteani et al., 2006; Rabiet et al., 2006). Therefore the in river waters in Pennsylvania and in Lake Erie, northeastern United States. Chem. Erde – Geochem. 66, 143–152. presence of anthropogenic Gd anomalies may also prove useful Bayer HealthCare AG, 2009. Magnevist Product Information. . REE may be a fast and cost-effective way to screen large sample Berliner Wasserbetriebe, 2008. [Electronic Article] Berliner Wasser. . sets in a short period of time. Berliner Wasserbetriebe, 2009. Wasser im Haushalt – alles unter einem Dach. . Biagi, B.A., Enyeart, J.J., 1990. Gadolinium blocks low-threshold and high-threshold calcium currents in pituitary-cells. Am. J. Physiol. 259, C515–C520. 4. Conclusions Biddau, R., Bensimon, M., Cidu, R., Parriaux, A., 2009. Rare earth elements in groundwater from different Alpine aquifers. Chem. Erde – Geochem. 69, 327– Anthropogenic Gd initially used as a contrast agent in magnetic 339. resonance imaging and known to occur as a microcontaminant in Birke, M., Rauch, U., Lorenz, H., 2009. Uranium in stream and mineral water of the Federal Republic of Germany. Environ. Geochem. Health 31, 693–706. rivers and groundwater could be traced in municipal tap water Broome, D.R., 2008. Nephrogenic systemic fibrosis associated with gadolinium used as drinking water in the City of Berlin, Germany. It was dem- based contrast agents: a summary of the medical literature reporting. Eur. J. onstrated that tap water in the western districts of the city shows Radiol. 66, 230–234. Cabella, C., Crich, S.G., Corpillo, D., Barge, A., Ghirelli, C., Bruno, E., Lorusso, V., Uggeri, large positive anthropogenic Gd anomalies, while tap water from F., Aime, S., 2006. Cellular labeling with Gd(III) chelates: only high the eastern districts does not. The anthropogenic Gd anomalies in thermodynamic stabilities prevent the cells acting as ‘sponges’ of Gd3+ ions. Contrast Media Mol. Imaging 1, 23–29. the western districts are as large as GdSN/Gd = 33.7, indicating SN Cowper, S.E., Robin, H.S., Steinberg, S.M., Su, L.D., Gupta, S., LeBoit, P.E., 2000. that up to 97% of total Gd is of anthropogenic origin. The maximum Scleromyxoedema-like cutaneous diseases in renal-dialysis patients. Lancet anthropogenic Gd concentration was found in a sample from the 356, 1000–1001. 147 S. Kulaksız, M. Bau / Applied Geochemistry 26 (2011) 1877–1885

Darrah, T.H., Prutsman-Pfeiffer, J.J., Poreda, R.J., Campbell, M.E., Hauschka, P.V., Möller, P., Paces, T., Dulski, P., Morteani, G., 2002. Anthropogenic Gd in surface Hannigan, R.E., 2009. Incorporation of excess gadolinium into human bone from water, drainage system, and the water supply of the city of Prague, Czech medical contrast agents. Metallomics 1, 479–488. Republic. Environ. Sci. Technol. 36, 2387–2394. deBoer, J.L.M., Verweij, W., vanderVeldeKoerts, T., Mennes, W., 1996. Levels of rare Möller, P., Morteani, G., Dulski, P., 2003. Anomalous gadolinium, cerium, and earth elements in Dutch drinking water and its sources. Determination by yttrium contents in the Adige and Isarco River waters and in the water of their inductively coupled plasma mass spectrometry and toxicological implications. tributaries (Provinces Trento and Bolzano/Bozen, NE Italy). Acta Hydrochim. A pilot study. Water Res. 30, 190–198. Hydrobiol. 31, 225–239. Elbaz-Poulichet, F., Seidel, J.-L., Othoniel, C., 2002. Occurrence of an anthropogenic Morteani, G., Möller, P., Fuganti, A., Paces, T., 2006. Input and fate of anthropogenic gadolinium anomaly in river and coastal waters of Southern France. Water Res. estrogens and gadolinium in surface water and sewage plants in the 36, 1102–1105. hydrological basin of Prague (Czech Republic). Environ. Geochem. Health 28, Evans, C.H., 1990. Biochemistry of the lanthanides. In: Frieden, E. (Ed.), Biochemistry 257–264. of the Elements. Plenum Press, New York, pp. 303–315. Nozaki, Y., Lerche, D., Alibo, D.S., Tsutsumi, M., 2000. Dissolved indium and rare Federal Institute of Risk Assessment, 2006. BfR korrigiert earth elements in three Japanese rivers and Tokyo Bay: evidence for Höchstmengenempfehlung für Uran in Wässern zur Zubereitung von anthropogenic Gd and In. Geochim. Cosmochim. Acta 64, 3975–3982. Säuglingsnahrung (Gemeinsame Stellungnahme Nr. 014/2006 des BfS und des Petelet-Giraud, E., Klaver, G., Negrel, P., 2009. Natural versus anthropogenic sources BfR vom 16.01.2006). . Hydrol. 369, 336–349. German Drinking Water Commission (TWK) of the Federal Ministry of Health, 2008. Rabiet, M., Brissaud, F., Seidel, J.-L., Pistre, S., Elbaz-Poulichet, F., 2005. Deciphering Uran im Trinkwasser. Stellungnahme der TWK zu sechs häufig gestellten the presence of wastewater in a medium-sized Mediterranean catchment using Fragen. Federal Environment Agency (UBA). . Rabiet, M., Togola, A., Brissaud, F., Seidel, J.L., Budzinski, H., Elbaz-Poulichet, F., 2006. Idée, J.M., Port, M., Raynal, I., Schaefer, M., Le Greneur, S., Corot, C., 2006. Clinical and Consequences of treated water recycling as regards pharmaceuticals and drugs biological consequences of transmetallation induced by contrast agents for in surface and ground waters of a medium-sized Mediterranean catchment. magnetic resonance imaging: a review. Fundam. Clin. Pharmacol. 20, Environ. Sci. Technol. 40, 5282–5288. 563–576. Rabiet, M., Brissaud, F., Seidel, J.L., Pistre, S., Elbaz-Poulichet, F., 2009. Positive Idee, J.M., Port, M., Dencausse, A., Lancelot, E., Corot, C., 2009. Involvement of gadolinium anomalies in wastewater treatment plant effluents and aquatic gadolinium chelates in the mechanism of nephrogenic systemic fibrosis: an environment in the Herault watershed (South France). Chemosphere 75, 1057– update. Radiol. Clin. North Am. 47, 855–869. 1064. Janssen, R.P.T., Verweij, W., 2003. Geochemistry of some rare earth elements in Schnug, E., Birke, M., Costa, N., Knolle, F., Panten, K., Lilienthal, H., Haneklaus, S., groundwater, Vierlingsbeek, The Netherlands. Water Res. 37, 1320–1350. 2008. Uranium in German tap and bottled waters. In: de Kok, L.J., Schnug, E.L. Knappe, A., Möller, P., Dulski, P., Pekdeger, A., 2005. Positive gadolinium anomaly in (Eds.), Loads and Fate of Fertilizer Derived Uranium. Backuys, Leiden, The surface water and ground water of the urban area Berlin, Germany. Chem. Erde Netherlands. – Geochem. 65, 167–189. Shabani, M.B., Akagi, T., Masuda, A., 1992. Preconcentration of trace rare-earth Kulaksız, S., Bau, M., 2007. Contrasting behaviour of anthropogenic gadolinium and elements in seawater by complexation with bis(2-ethylhexyl) hydrogen natural rare earth elements in estuaries and the gadolinium input into the phosphate and 2-ethylhexyl dihydrogen phosphate adsorbed on a C18 North Sea. Earth Planet. Sci. Lett. 260, 361–371. cartridge and determination by inductively coupled plasma mass Kümmerer, K., Helmers, E., 2000. Hospital effluents as a source of gadolinium in the spectrometry. Anal. Chem. 64, 737–743. aquatic environment. Environ. Sci. Technol. 34, 573–577. Shellock, F.G., Spinazzi, A., 2008. MRI Safety Update 2008: Part I, MRI contrast Künnemeyer, J., Terborg, L., Meermann, B., Brauckmann, C., Scheffer, I.M.A., Karst, U., agents and Nephrogenic Systemic Fibrosis. Am. J. Roentgenol. 191, 1129–1139. 2009. Speciation analysis of gadolinium chelates in hospital effluents and Thakral, C., Alhariri, J., Abraham, J.L., 2007. Long-term retention of gadolinium in wastewater treatment plant sewage by a Novel HILIC/ICP-MS method. Environ. tissues from nephrogenic systemic fibrosis patient after multiple gadolinium- Sci. Technol. 43, 2884–2890. enhanced MRI scans: case report and implications. Contrast Media Mol. Imaging Lawrence, M.G., 2010. Detection of anthropogenic gadolinium in the Brisbane 2 (4), 199–205. River plume in Moreton Bay, Queensland, Australia. Mar. Pollut. Bull. 60, Tricca, A., Stille, P., Steinmann, M., Kiefel, B., Samuel, J., Eikenberg, J., 1999. Rare 1113–1116. earth elements and Sr and Nd isotopic compositions of dissolved and suspended Lawrence, M.G., Jupiter, S.D., Kamber, B.S., 2006. Aquatic geochemistry of the rare loads from small river systems in the Vosges mountains (France), the river earth elements and yttrium in the Pioneer River catchment, Australia. Mar. Rhine and groundwater. Chem. Geol. 160, 139–158. Freshwater Res. 57, 725–736. Verplanck, P.L., Taylor, H.E., Nordstrom, D.K., Barber, L.B., 2005. Aqueous stability of Lawrence, M.G., Ort, C., Keller, J., 2009. Detection of anthropogenic gadolinium in gadolinium in surface waters receiving sewage treatment plant effluent, treated wastewater in South East Queensland, Australia. Water Res. 43, 3534– Boulder Creek, Colorado. Environ. Sci. Technol. 39, 6923–6929. 3540. Verplanck, P.L., Furlong, E.T., Gray, J.L., Phillips, P.J., Wolf, R.E., Esposito, K., 2010. Lawrence, M.G., Keller, J., Poussade, Y., 2010. Removal of magnetic resonance Evaluating the behavior of gadolinium and other rare earth elements through imaging contrast agents through advanced water treatment plants. Water Sci. large metropolitan sewage treatment plants. Environ. Sci. Technol. 44, 3876– Technol. 61, 685–692. 3882. Massmann, G., Sültenfuss, J., Dünnbier, U., Knappe, A., Taute, T., Pekdeg˘er, A., 2008. World Health Organization, 2008. Guidelines for Drinking-water Quality [electronic Investigation of groundwater residence times during bank filtration in Berlin: resource]: Incorporating 1st and 2nd Addenda. Recommendations, third ed., vol. multi-tracer approach. Hydrol. Process. 22, 788–801. 1. World Health Organization, Geneva. McLennan, S.M., 1989. Rare earth elements in sedimentary rocks; influence of Yang, X.C., Sachs, F., 1989. Block of stretch-activated ion channels in xenopus provenance and sedimentary processes. Rev. Mineral. Geochem. 21, 169–200. oocytes by gadolinium and calcium-ions. Science 243, 1068–1071. Molgo, J., Delpozo, E., Banos, J.E., Angautpetit, D., 1991. Changes of quantal Yantasee, W., Fryxell, G.E., Porter, G.A., Pattamakomsan, K., Sukwarotwat, V., transmitter release caused by gadolinium ions at the frog neuromuscular- Chouyyok, W., Koonsiripaiboon, V., Xu, J., Raymond, K.N., 2010. Novel sorbents junction. Brit. J. Pharmacol. 104, 133–138. for removal of gadolinium-based contrast agents in sorbent dialysis and Möller, P., Dulski, P., 2010a. Transmetallation of Gd-DTPA by Cu, Y and lanthanides hemoperfusion: preventive approaches to nephrogenic systemic fibrosis. and its impact on the hydrosphere. Appl. Geochem. 25, 48–59. Nanomed. – Nanotechnol. Biol. Med. 6, 1–8. Möller, P., Dulski, P., 2010b. Gd-DTPA in the hydrosphere: kinetics of Zhu, Y., Hoshino, M., Yamada, H., Itoh, A., Haraguchi, H., 2004. Gadolinium anomaly transmetallation by ions of rare earth elements. Chem. Erde – Geochem. 70, in the distributions of rare earth elements observed for coastal seawater and 125–136. river waters around Nagoya City. Bull. Chem. Soc. Jpn. 77, 1835–1842. Möller, P., Dulski, P., Bau, M., Knappe, A., Pekdeger, A., Sommer-von Jarmersted, C., Zhu, Y., Hattori, R., Rahmi, D., Itoh, S.O., Fujimori, E., Umemura, T., Haraguchi, H., 2000. Anthropogenic gadolinium as a conservative tracer in hydrology. J. 2005. Fractional distributions of trace metals in surface water of Lake Biwa as Geochem. Explor. 69–70, 409–414. studied by ultrafiltration and ICP-MS. Bull. Chem. Soc. Jpn. 78, 1970–1976.

148 149 150 CHAPTER VII – CONCLUSIONS

The first comprehensive geochemical dataset of the Rhine River and its major tributaries is provided with this study. While the presence of anthropogenic Gd had been reported previously, the presence of anthropogenic La and anthropogenic Sm has been documented for the first time, both for the Rhine River and natural waters worldwide.

1. DOCUMENTING AND QUANTIFYING ANTHROPOGENIC RARE EARTH ELEMENTS

1.1 ANTHROPOGENIC GD

The presence of anthropogenic Gd in the majority of the tributaries of the Rhine River is documented for the first time, especially for the less studied tributaries of the High Rhine and Upper Rhine. Further downstream, within the Middle Rhine and Lower Rhine sections, the presence of anthropogenic Gd in the Rhine River and its tributaries has been confirmed. Over the past 15 years (Bau and Dulski, 1996), Gd anomalies have increased up to 2‐fold in the Wupper River, 5‐ to 15‐fold in the Sieg River and 2‐ to 8‐fold in the Rhine River south of

* Düsseldorf. Anthropogenic Gd anomalies (GdSN/Gd SN) of up to 62 are observed in the High Rhine, corresponding to an anthropogenic Gd concentration of 50.4 ng/kg on top of 0.8 ng/kg of natural Gd in the filtrate. Among the tributaries, Lippe River and Wupper River show the highest

* anthropogenic Gd (GdSN/Gd SN = 219 and 64). Surprisingly, some of the tributaries of the High * Rhine also exhibit large Gd anomalies (e.g. Töss River, GdSN/Gd SN = 65). Additionally, Lake Constance, under protection and used as a drinking water source, also shows a small but

* significant Gd anomaly (GdSN/Gd SN = 2.1).

1.2 ANTHROPOGENIC LA

The presence of anthropogenic La in the Rhine River downstream from Rhine‐km 447.3 is

* documented for the first time. Anthropogenic La anomalies (LaSN/La SN) of up to 46.3 are observed in the Rhine River, corresponding to 331 ng/kg of anthropogenic La on top of 7.3 ng/kg of natural La in the dissolved filtrate. In contrast, samples from upstream of this location do not show anthropogenic La, and none of the tributaries exhibit significant La anomalies. The source of the contamination is traced to the effluent pipe of a fluid catalytic cracking (FCC) catalyst

151 producing plant at Rhine‐km 447.3. The highest concentration of La measured in the vicinity of this outlet is 49 mg/kg.

1.3 ANTHROPOGENIC SM

The presence of anthropogenic Sm in the Rhine River is documented for the first time. Anthropogenic Sm anomalies as high as 7.7 have been observed in the Rhine River at Leverkusen, Germany, corresponding to an anthropogenic Sm concentration of 7.3 ng/kg on top of a natural Sm concentration of 1.1 ng/kg. Preliminary data indicates that the source of Sm contamination is the same point‐source as that for La contamination: effluent pipe of a FCC catalyst producing plant at Rhine‐km 447.3. Anthropogenic Sm anomalies are not observed in samples prior to October 2010 and have emerged some time between then and May 2011. As the discovery of anthropogenic Sm anomalies was not predicted and is more recent, it has not been a part of the original goals of this study. Hence, anthropogenic Sm has not been mentioned explicitly in the objectives section.

2. CONTROLS ON RARE EARTH ELEMENTS IN THE RHINE RIVER CATCHMENT

First comprehensive dataset of REE for the Rhine River is reported, starting from Lake Constance and the border between Switzerland and Germany, and ending close to the border between Germany and the Netherlands. This information does not only serve to quantify anthropogenic REE anomalies, but will also provide background levels for future studies.

2.1 NATURAL RARE EARTH ELEMENTS

With only a few exceptions, shale‐normalized REE patterns of the Rhine River are slightly to significantly LREE depleted, with an average NdSN/YbSN ratio of 0.23 and a range between 0.05 and 0.43. Most of the tributaries are also LREE depleted (0.04 < NdSN/YbSN < 0.63, average 0.23) showing similar shale‐normalized REE patterns. A number of controls have been identified on the natural and anthropogenic dissolved REE concentrations and distribution in the Rhine River and its tributaries. On average pH of the Rhine River is lower in the low discharge season. Since REE concentrations generally increase with decreasing pH, lower pH during lower discharge has the combined effect of relatively small differences in REE concentrations between sampling

152 campaigns. Discharge plays an important role in the Rhine River, increasing dissolved REE concentrations due to the increase of colloids as discharge increases within each sampling campaign. Additionally, large‐scale changes in catchment geology affect REE distribution in the Rhine River and its tributaries. The transition from the Upper Rhine dominated by sandstones and siltstones into the Middle Rhine dominated by claystones and shales causes 2‐ to 3‐fold increase in REE concentrations and 20 – 30% increase in NdSN/YbSN ratios.

2.2 ANTHROPOGENIC RARE EARTH ELEMENTS

Anthropogenic REE are also affected by changes in discharge in two different ways. At the same sampling location, an increase in discharge causes an increase in natural REE, decreasing normalized ratio of anthropogenic REE to natural (background) REE. Furthermore, as the input of anthropogenic REE is not affected by discharge, an increase in discharge at the same sampling location causes dilution of the anthropogenic REE, effectively decreasing normalized ratio of anthropogenic REE to natural (background) REE.

Anthropogenic La decreases with increasing discharge within each sampling campaign, while the opposite is true for anthropogenic Gd. With increasing discharge in the downstream direction of the Rhine River, anthropogenic Gd anomalies also increase. This is due to the mounting input from WWTP effluents with increasing Rhine‐km that exceeds the effect of dilution due to the increase in discharge downstream. The reason for this increase in the input of anthropogenic Gd is the rise in population density downstream of the river. The rivers in the Middle Rhine and Lower Rhine show the highest anthropogenic Gd anomalies, while also draining areas with high population density.

3. THE ROLE OF ULTRAFILTRATION

Ultrafiltration of Rhine River water from Leverkusen provides valuable knowledge on the partitioning of natural and anthropogenic REE between the nanoparticulate (< 0.2 µm) and truly dissolved (between 10 kDa and 0.2 µm) REE pools. Removal of natural REE by ultrafiltration is most significant for the LREE and decreases systematically with increasing atomic number. While 83% of Ce is removed from solution via ultrafiltration, only 30% of Yb and Lu are removed, i.e.

153 ultrafiltration increases the HREE enrichment in the solution. This finding is in agreement with previous results that show a HREE‐enriched truly dissolved pool and a nanoparticulate pool that is dominated by the more particle‐reactive LREE.

Ultrafiltration also reveals that anthropogenic La and Sm are present in both the truly dissolved and nanoparticulate pools. In contrast, anthropogenic Gd is exclusively present in the truly dissolved pool and there are no Gd anomalies in the nanoparticulate pool.

4. RARE EARTH ELEMENTS AS EMERGING CONTAMINANTS IN TAP WATER

1 6

0.1

REE / PAAS x 10 /PAAS REE Pankow (Berlin - GER) Reichstag (Berlin - GER) Köln (GER) Düsseldorf (GER) London (GBR) 0.01

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 1. Shale normalized REE patterns for tap water from selected cities.

Although the existence of Gd anomalies in tap water had been reported to some degree, a comprehensive study covering different parts of a city had been missing in literature. In order to fill this gap in the transport pathways of anthropogenic Gd, tap water from all boroughs of Berlin was analyzed. Berlin has been an excellent choice for a study of this kind, as water management strategies are markedly different in both former east and west parts of the city. The west part

154 relies heavily on natural and artificial groundwater recharge while the east part does not. This is clearly reflected in the results of this study (Chapter VI) with a remarkable difference in the anthropogenic Gd content of Berlin tap water between the east and west boroughs. The largest anthropogenic Gd anomalies were found in the former western parts (Fig. 1 ‐ sample from the

* former parliament building Reichstag: GdSN/Gd SN = 34), while samples from the former eastern * boroughs showed no anthropogenic Gd anomalies (GdSN/Gd SN < 1.60). These results have significantly consolidated the concept of anthropogenic Gd as a tracer in hydrogeology. Pharmaceuticals and other highly stable compounds are more likely to be found in tap water that shows an anthropogenic Gd anomaly, than tap water that does not.

A large percentage of the population that lives in the vicinity of the Rhine River relies to a great extent on Rhine River water as a source of drinking water. Preliminary data from Köln and Düsseldorf (Fig. 1) shows moderate anthropogenic Gd anomalies in both cities as well as a small anthropogenic La anomaly. The population density in the area of these two cities is rather high, and both cities rely heavily on bank infiltration of Rhine River water for groundwater recharge. Additionally, tap water from London (U.K.) also carries a small but significant anthropogenic Gd anomaly. In this case, water from the Thames River is used for replenishing groundwater levels. Future work should include sampling tap water from major cities of Europe that directly or indirectly rely on surface waters for drinking water. Establishing a database of anthropogenic REE in tap water will be an important addition to studies of pharmaceuticals and emerging contaminants in drinking water.

155 5. CLOSING REMARKS

With the addition of Sm to the already heavily burdened anthropogenic load of REE in the Rhine River, shale normalized REE patterns have become exceedingly peculiar. In parts of the Rhine River, anthropogenic REE concentrations are an order of magnitude higher than background concentrations for three (La, Sm and Gd) of the fourteen REE measured. Due to the high toxicity of REE, this may pose a threat to the ecosystem, especially given the lack of studies that focus on low‐dosage long‐term exposure. While the existence of ecotoxicological effects remains to be shown in the future, the immediate effect on natural REE studies is unquestionable.

With the demand for REE and the number of REE applications rapidly increasing, more anthropogenic REE are expected to enter the environment in the near future. Thus, the fate and transport of anthropogenic REE should be studied in depth, including any potential (and synergetic) toxicity effects, while distribution and behavior of the natural REE in different compartments of the environment should be established and closely monitored. Establishing REE background concentrations in industrialized countries will soon no longer be possible due to widespread contamination, also placing future REE studies at risk. Studies of emerging contaminants are already underway and anthropogenic REE will be one of several key areas to study in the future. The increasing use of formerly exotic high‐tech metals is not confined to the REE, and the current situation of REE contamination may well be only a prelude to what will soon happen with other emerging contaminants.

156 157 158

Appendix 1 – Shale‐normalized REE patterns of the Rhine River

100 100 Rhine 34.4 km - May 08 Rhine/Thur 65.1 km - May 08 Rhine 34.4 km - May 09 Rhine/Thur 65.1 km - May 09 Rhine 34.4 km - Oct 09 Rhine/Thur 65.1 km - Oct 09 Lake Constance -38 km - Aug 11 Lake Constance -38 km - Aug 11

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

1 2

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Töss 73.3 km - May 08 Rhine/Wutach 100.5 km - May 08 Rhine/Töss 73.3 km - May 09 Rhine/Wutach 100.5 km - May 09 Rhine/Töss 73.3 km - Oct 09 Rhine/Wutach 100.5 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

3 4

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

159

Appendix 1 continued.

100 100 Rhine/Aare 103 km - May 08 Rhine/Birs 154 km - May 08 Rhine/Aare 103 km - May 09 Rhine/Aare 103 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

5 6

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Wiese 170.6 km - May 08 Rhine 210.5 km - May 08 Rhine/Wiese 170.6 km - May 09 Rhine/Wiese 170.6 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

7 8

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

160

Appendix 1 continued.

100 100 Rhine/Elz 260.9 km - May 08 Rhine/Kinzig 294.2 km - May 08 Rhine/Elz 260.9 km - May 09 Rhine/Kinzig 294.2 km - May 09 Rhine/Elz 260.9 km - Oct 09 Rhine/Kinzig 294.2 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

9 10

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Ill 304.6 km - May 08 Rhine/Acher/Rench 313.2 km - May 08 Rhine/Ill 304.6 km - May 09 Rhine/Ill 304.6 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

11 12

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

161

Appendix 1 continued.

100 100 Rhine 318.5 km - May 08 Rhine/Murg 340.4 km - May 08

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

13 14

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine 354.2 km - May 08 Rhine 372.1 km - May 08 Rhine 354.3 km - May 09 Rhine 372 km - May 09 Rhine 354.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

15 16

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

162

Appendix 1 continued.

100 100 Rhine 393.9 km - May 08 Rhine 414.3 km - May 08 Rhine 414.3 km - May 09 Rhine 414.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

17 18

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Pfrimm 446.7 km - May 09 Rhine/Main 493 km - May 08 Rhine/Pfrimm 446.7 km - Oct 09 Rhine/Main 493 km - May 09 Rhine/Main 493 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

19 20

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

163

Appendix 1 continued.

100 100 Rhine/Nahe 525.6 km - May 08 Rhine/Lahn 584.5 km - May 08 Rhine/Nahe 525.6 km - May 09 Rhine/Lahn 584.5 km - May 09 Rhine/Nahe 525.6 km - Oct 09 Rhine/Lahn 584.5 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

21 22

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Mosel 591.5 km - May 08 Rhine/Ahr 627.7 km - May 08 Rhine/Mosel 591.5 km - Oct 09 Rhine/Ahr 627.7 km - May 09 Rhine/Ahr 627.7 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

Sample not included in Rhine-km plots 23 24

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

164

Appendix 1 continued.

100 100 Rhine/Sieg 655.1 km - May 08 Rhine/Wupper 703 km - May 08 Rhine/Sieg 655.1 km - May 09 Rhine/Wupper 703 km - May 09 Rhine/Sieg 655.1 km - Oct 09 Rhine/Wupper 703 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

25 26

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine/Erft 735.4 km - May 08 Rhine/Ruhr 779.4 km - May 08 Rhine/Erft 735.4 km - May 09 Rhine/Ruhr 779.4 km - May 09 Rhine/Erft 735.4 km - Oct 09 Rhine/Ruhr 779.4 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

27 28

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

165

Appendix 1 continued.

100 100 Rhine/Lippe 811.4 km - May 08 Rhine 823.3 km - May 09 Rhine/Lippe 811.4 km - May 09 Rhine 823.3 km - May 09 Rhine/Lippe 811.4 km - Oct 09 Rhine 823.3 km - May 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

29 30

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Rhine 823.3 km - Oct 09 Rhine 857.3 km - May 08 Rhine 823.3 km - Oct 09 Rhine 823.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

31 32

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

166

Appendix 2 – Shale‐normalized REE patterns of the tributaries to the Rhine River

100 100 Thur 66.6 km - May 08 Töss 73.4 km - May 08 Thur 66.6 km - May 09 Töss 73.4 km - May 09 Thur 66.6 km - Oct 09 Töss 73.4 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

1 2

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Wutach 101 km - May 08 Aare 103.2 km - May 08 Wutach 101 km - May 09 Aare 103.2 km - May 09 Wutach 101 km - Oct 09 Aare 103.2 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

3 4

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

167

Appendix 2 continued.

100 100 Birs 164 km - May 08 Wiese 169.2 km - May 08 Wiese 169.2 km - May 09 Wiese 169.2 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

5 6

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Elz 267.5 km - May 08 Kinzig 299.4 km - May 08 Elz 267.5 km - May 09 Kinzig 299.4 km - May 09 Elz 267.5 km - Oct 09 Kinzig 299.4 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

7 8

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

168

Appendix 2 continued.

100 100 Ill 312 km - May 08 Acher/Rench 315.3 km - May 08 Ill 312 km - May 09 Acher/Rench 315.3 km - May 09 Ill 312 km - Oct 09 Acher/Rench 315.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

9 10

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Moder 334.4 km - May 08 Sauer 344 km - May 09 Moder 334.4 km - May 09 Sauer 344 km - Oct 09 Moder 334.4 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

11 12

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

169

Appendix 2 continued.

100 100 Murg 344.6 km - May 08 Lauter 355.6 km - May 09 Murg 344.6 km - May 09 Lauter 355.6 km - Oct 09 Murg 344.6 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

13 14

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Neckar 428.2 km - May 08 Pfrimm 446.7 km - May 09 Neckar 428.2 km - May 09 Pfrimm 446.7 km - Oct 09 Neckar 428.2 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

15 16

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

170

Appendix 2 continued.

100 100 Main 496.7 km - May 08 Nahe 529.2 km - May 08 Main 496.7 km - May 09 Nahe 529.2 km - May 09 Main 496.7 km - Oct 09 Nahe 529.2 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

17 18

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Lahn 585.8 km - May 08 Mosel 592.3 km - May 08 Lahn 585.8 km - May 09 Mosel 592.3 km - May 09 Lahn 585.8 km - Oct 09 Mosel 592.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

19 20

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

171

Appendix 2 continued.

100 100 Wied 610.3 km - May 08 Ahr 629.3 km - May 08 Wied 610.3 km - May 09 Ahr 629.3 km - May 09 Wied 610.3 km - Oct 09 Ahr 629.3 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

21 22

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu 100 100 Sieg 659.5 km - May 08 Wupper 703.4 km - May 08 Sieg 659.5 km - May 09 Wupper 703.4 km - May 09 Sieg 659.5 km - Oct 09 Wupper 703.4 km - May 09 Wupper 703.4 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 PAAS / REE x 10 PAAS / REE

0.1 0.1

23 24

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

172

Appendix 2 continued.

100 100 Erft 735.6 km - May 08 Ruhr 780.7 km - May 08 Erft 735.6 km - May 09 Ruhr 780.7 km - May 08 Erft 735.6 km - Oct 09 Ruhr 780.7 km - May 09 Ruhr 780.7 km - Oct 09

10 10 6 6

1 1 REE / PAAS x 10 REE / PAAS x 10

0.1 0.1

25 26

0.01 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

100

10 6

1 REE / PAAS x 10 PAAS / REE

0.1

Lippe 814.6 km - May 08 Lippe 814.6 km - May 09 27 Lippe 814.6 km - Oct 09 0.01 La Ce Pr NdPmSm Eu Gd Tb Dy Ho Er Tm Yb Lu

173

Appendix 3 – May 2008: sample details and concentrations for the Rhine River

Sample Rhine Rhine/Thur Rhine/Töss Rhine/Wutach Rhine/Aare Rhine/Birs Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 34.4 65.1 73.3 100.5 103 154 Date 24/05/2008 25/05/2008 25/05/2008 25/05/2008 25/05/2008 26/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Sat Sun Sun Sun Sun Mon Time 17:00 12:10 13:10 15:10 15:40 11:20 Latitude 47.69078 47.60646 47.55163 47.61611 47.60603 47.5432 Longitude 8.75336 8.59811 8.55629 8.25751 8.22718 7.72479 Temperature oC 16 16 16.4 16.9 17 17.8 pH 8.44 8.3 8.33 8.31 8.38 8.27 EC µS/cm 336 344 350 355 361 336 discharge m3/s 426 ‐ ‐ ‐ 1060 ‐ Na+ mg/kg ‐ ‐‐ ‐‐ ‐ Mg2+ mg/kg ‐ ‐‐ ‐‐ ‐ Si mg/kg ‐ ‐‐ ‐‐ ‐ Cl‐ mg/kg ‐ ‐‐ ‐‐ ‐ K+ mg/kg ‐ ‐‐ ‐‐ ‐ Ca2+ mg/kg ‐ ‐‐ ‐‐ ‐ ‐ HCO3 mg/kg ‐ ‐‐ ‐‐ ‐ ‐ NO3 mg/kg ‐ ‐‐ ‐‐ ‐ Br‐ mg/kg ‐ ‐‐ ‐‐ ‐ 2‐ SO4 mg/kg ‐ ‐‐ ‐‐ ‐ Rb µg/kg 1.04 1.01 1.05 1.1 1.15 1.2 Sr mg/kg 0.39 0.39 0.37 0.38 0.39 0.32 Y ng/kg 8.74 10.2 9.13 9.3 9.4 8.11 Ba µg/kg 29 30.2 30 30.5 32.4 28.4 U µg/kg 1.09 1.08 0.99 1.02 1.04 0.76 La ng/kg 1.85 1.91 1.01 0.97 1 5.28 Ce ng/kg 2.68 3.03 1.35 1.29 1.41 7.71 Pr ng/kg 0.62 0.67 0.37 0.39 0.39 0.85 Nd ng/kg 3.14 3.02 2.29 2.36 2.4 3.14 Sm ng/kg 0.9 0.92 0.88 0.88 0.9 0.79 Eu ng/kg 0.26 0.25 0.24 0.23 0.25 ‐ Gd ng/kg 2.57 2.5 2.45 2.7 2.57 2.13 Tb ng/kg 0.22 0.21 0.21 0.24 0.22 0.17 Dy ng/kg 1.41 1.41 1.46 1.49 1.48 1.08 Ho ng/kg 0.31 0.3 0.3 0.31 0.33 0.23 Er ng/kg 0.96 0.92 0.94 0.94 0.95 0.71 Tm* ng/kg 0.16 0.13 0.14 0.13 0.14 0.12 Yb ng/kg 1.25 0.95 0.97 0.93 0.95 0.96 Lu ng/kg 0.21 0.17 0.17 0.17 0.19 0.24

Laanthr ng/kg 0.32 0.57 0.29 0.2 0.22 3.41

Lanat ng/kg 1.54 1.34 0.72 0.77 0.78 1.87

Gdanthr ng/kg 1.26 1.06 0.71 1.01 0.83 1.12

Gdnat ng/kg 1.31 1.44 1.73 1.69 1.74 1.01 ΣREE ng/kg 16.5 16.4 12.8 13 13.2 23.4 ΣREE (Natural) ng/kg 14.9 14.8 11.8 11.8 12.1 18.9 (La + Gd) / ΣREE % 9.6 9.9 7.9 9.3 7.9 19 * LaSN / La SN 1.21 1.42 1.4 1.26 1.28 2.82 * CeSN / Ce SN 0.63 0.8 0.59 0.53 0.57 1.6 * EuSN / Eu SN 1.12 1.02 0.9 0.88 0.94 ‐ * GdSN / Gd SN 1.96 1.74 1.41 1.6 1.47 2.11

NdSN/YbSN 0.21 0.26 0.2 0.21 0.21 0.27

TbSN/YbSN 0.63 0.80 0.80 0.94 0.83 0.63

174

Appendix 3 continued.

Sample Rhine/Wiese Rhine Rhine/Elz Rhine/Kinzig Rhine/Ill Rhine/Acher/Rench Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 170.6 210.5 260.9 294.2 304.6 313.2 Date 26/05/2008 26/05/2008 26/05/2008 26/05/2008 27/05/2008 27/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Mon Mon Mon Mon Tue Tue Time 13:30 14:25 15:20 16:35 09:25 10:55 Latitude 47.59248 47.90797 48.31209 48.57306 48.65078 48.70783 Longitude 7.59193 7.58136 7.71693 7.803647 7.86636 7.94897 Temperature oC 16.7 18.8 18.1 17.7 17.9 18.4 pH 8.25 8.37 8.22 8.12 7.97 8.19 EC µS/cm 330 390 379 391 367 378 discharge m3/s 1180 ‐ 914 1080 1250 1240 Na+ mg/kg ‐‐‐ ‐‐ ‐ Mg2+ mg/kg ‐‐‐ ‐‐ ‐ Si mg/kg ‐‐‐ ‐‐ ‐ Cl‐ mg/kg ‐‐‐ ‐‐ ‐ K+ mg/kg ‐‐‐ ‐‐ ‐ Ca2+ mg/kg ‐‐‐ ‐‐ ‐ ‐ HCO3 mg/kg ‐‐‐ ‐‐ ‐ ‐ NO3 mg/kg ‐‐‐ ‐‐ ‐ Br‐ mg/kg ‐‐‐ ‐‐ ‐ 2‐ SO4 mg/kg ‐‐‐ ‐‐ ‐ Rb µg/kg 1.44 1.45 1.35 1.35 1.62 1.86 Sr mg/kg 0.3 0.34 0.32 0.33 0.31 0.4 Y ng/kg 9.78 6.99 9.75 10.4 16.2 14.8 Ba µg/kg 35.2 47.8 34.3 35.8 39 43.5 U µg/kg 0.69 0.71 0.74 0.74 0.71 0.76 La ng/kg 2.49 1.34 1.68 2.1 2.22 2.69 Ce ng/kg 3.1 2.11 2.06 2.58 2.78 3.58 Pr ng/kg 0.72 0.43 0.48 0.55 0.57 0.68 Nd ng/kg 3.51 2.53 2.54 2.8 2.8 3.4 Sm ng/kg 0.97 0.73 0.79 0.76 0.81 1.05 Eu ng/kg 0.26 0.19 0.21 0.22 0.22 0.28 Gd ng/kg 2.26 1.95 3.33 3.11 3.06 2.99 Tb ng/kg 0.2 0.15 0.19 0.19 0.18 0.22 Dy ng/kg 1.36 1.02 1.25 1.24 1.19 1.47 Ho ng/kg 0.3 0.23 0.28 0.28 0.27 0.34 Er ng/kg 0.96 0.72 0.88 0.92 0.88 1.16 Tm* ng/kg 0.15 0.12 0.13 0.14 0.13 0.18 Yb ng/kg 1.14 1.02 1.01 1.06 1.03 1.38 Lu ng/kg 0.23 0.21 0.18 0.2 0.21 0.27

Laanthr ng/kg 0.68 0.13 0.57 0.63 0.88 1.21

Lanat ng/kg 1.81 1.22 1.11 1.47 1.33 1.48

Gdanthr ng/kg 0.9 0.87 2.09 2.05 1.85 1.33

Gdnat ng/kg 1.37 1.08 1.24 1.07 1.21 1.66 ΣREE ng/kg 17.7 12.8 15 16.2 16.3 19.7 ΣREE (Natural) ng/kg 16.1 11.8 12.3 13.5 13.6 17.2 (La + Gd) / ΣREE % 8.9 7.8 18 17 17 13 * LaSN / La SN 1.38 1.1 1.51 1.43 1.66 1.82 * CeSN / Ce SN 0.63 0.63 0.65 0.65 0.75 0.85 * EuSN / Eu SN 1.05 0.99 1.02 1.12 1.02 0.99 * GdSN / Gd SN 1.66 1.8 2.68 2.92 2.54 1.8

NdSN/YbSN 0.26 0.21 0.21 0.22 0.23 0.2

TbSN/YbSN 0.64 0.54 0.70 0.65 0.64 0.58

175

Appendix 3 continued.

Sample Rhine Rhine/Murg Rhine Rhine Rhine Rhine Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 318.5 340.4 354.2 372.1 393.9 414.3 Date 27/05/2008 27/05/2008 27/05/2008 27/05/2008 27/05/2008 28/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Tue Tue Tue Tue Tue Wed Time 11:25 13:30 14:35 15:25 16:10 09:50 Latitude 48.75402 48.88728 48.97693 49.11576 49.28316 49.43374 Longitude 7.97143 8.13714 8.25658 8.36573 8.47361 8.50629 Temperature oC 18.8 18.5 18.8 19 20.2 18.9 pH 8.17 8.15 8.07 8.1 8.14 8.08 EC µS/cm 375 390 391 398 405 382 discharge m3/s 1230 1310 1310 1310 1290 1280 Na+ mg/kg ‐ ‐‐‐‐ ‐ Mg2+ mg/kg ‐ ‐‐‐‐ ‐ Si mg/kg ‐ ‐‐‐‐ ‐ Cl‐ mg/kg ‐ ‐‐‐‐ ‐ K+ mg/kg ‐ ‐‐‐‐ ‐ Ca2+ mg/kg ‐ ‐‐‐‐ ‐ ‐ HCO3 mg/kg ‐ ‐‐‐‐ ‐ ‐ NO3 mg/kg ‐ ‐‐‐‐ ‐ Br‐ mg/kg ‐ ‐‐‐‐ ‐ 2‐ SO4 mg/kg ‐ ‐‐‐‐ ‐ Rb µg/kg 1.78 1.79 1.79 1.58 1.66 1.83 Sr mg/kg 0.39 0.38 0.35 0.3 0.3 0.36 Y ng/kg 14.5 16.6 13.3 17.1 14.8 17.5 Ba µg/kg 40.2 42.4 38.8 38.5 39.9 44.8 U µg/kg 0.74 0.75 0.87 0.81 0.81 0.73 La ng/kg 8.5 2.37 2.17 4.12 2.43 2.74 Ce ng/kg 3.35 2.71 2.25 5.25 2.62 3.71 Pr ng/kg 0.86 0.81 0.67 1.64 0.95 1.04 Nd ng/kg 3.61 4.17 3.51 8.81 5.18 5.4 Sm ng/kg 1.04 1.3 1.09 2.27 1.54 1.49 Eu ng/kg 0.28 0.33 0.29 0.61 0.44 0.4 Gd ng/kg 2.98 3.62 3.61 5.19 4.19 4.43 Tb ng/kg 0.22 0.28 0.25 0.42 0.31 0.32 Dy ng/kg 1.46 1.82 1.7 2.55 2.11 2.05 Ho ng/kg 0.34 0.41 0.38 0.55 0.44 0.48 Er ng/kg 1.09 1.31 1.21 1.71 1.45 1.48 Tm* ng/kg 0.16 0.19 0.19 0.25 0.21 0.22 Yb ng/kg 1.16 1.35 1.4 1.78 1.54 1.55 Lu ng/kg 0.2 0.22 0.26 0.28 0.27 0.29

Laanthr ng/kg 6.77 0.58 0.65 ‐0.9 0.05

Lanat ng/kg 1.74 1.78 1.52 5.04 2.38 2.77

Gdanthr ng/kg 1.44 1.54 1.89 2.2 1.83 2.31

Gdnat ng/kg 1.54 2.08 1.72 2.99 2.36 2.12 ΣREE ng/kg 25.3 20.9 19 35.4 23.7 25.6 ΣREE (Natural) ng/kg 17 18.7 16.4 34.1 21.8 23.3 (La + Gd) / ΣREE % 33 10 13 3.6 7.9 8.9 * LaSN / La SN 4.9 1.33 1.43 0.82 1.02 0.99 * CeSN / Ce SN 0.7 0.53 0.52 0.4 0.39 0.49 * EuSN / Eu SN 1.05 0.94 0.99 1.11 1.1 1.07 * GdSN / Gd SN 1.94 1.74 2.1 1.74 1.77 2.09

NdSN/YbSN 0.26 0.26 0.21 0.41 0.28 0.29

TbSN/YbSN 0.70 0.74 0.65 0.86 0.74 0.76

176

Appendix 3 continued.

Sample Rhine/Main Rhine/Nahe Rhine/Lahn Rhine/Mosel Rhine/Ahr Rhine/Sieg Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 493 525.6 584.5 591.5 627.7 655.1 Date 28/05/2008 28/05/2008 28/05/2008 28/05/2008 29/05/2008 29/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Wed Wed Wed Wed Thu Thu Time 12:10 13:50 17:15 15:50 10:15 11:45 Latitude 49.97127 49.96998 50.3004 50.35874 50.54577 50.74 Longitude 8.32867 7.93861 7.60238 7.60476 7.28068 7.11326 Temperature oC 20 21.4 21.3 21.2 21.4 19.9 pH 7.99 8.04 8.12 8.12 8.09 8.06 EC µS/cm 460 479 508 501 509 525 discharge m3/s 1380 1600 1580 1600 1770 1740 Na+ mg/kg ‐‐‐‐‐ ‐ Mg2+ mg/kg ‐‐‐‐‐ ‐ Si mg/kg ‐‐‐‐‐ ‐ Cl‐ mg/kg ‐‐‐‐‐ ‐ K+ mg/kg ‐‐‐‐‐ ‐ Ca2+ mg/kg ‐‐‐‐‐ ‐ ‐ HCO3 mg/kg ‐‐‐‐‐ ‐ ‐ NO3 mg/kg ‐‐‐‐‐ ‐ Br‐ mg/kg ‐‐‐‐‐ ‐ 2‐ SO4 mg/kg ‐‐‐‐‐ ‐ Rb µg/kg 2.19 1.88 2.28 1.95 2.55 2.58 Sr mg/kg 0.35 0.31 0.33 0.31 0.36 0.37 Y ng/kg 22.1 22.2 61.7 28.6 37.9 37.8 Ba µg/kg 57.9 55 47.3 63.9 50.6 49.5 U µg/kg 0.72 0.72 0.83 0.71 0.89 0.96 La ng/kg 45.1 117 89.1 84.9 90.2 89.3 Ce ng/kg 4.49 6.36 7.97 8.38 10 8.9 Pr ng/kg 1.95 3.17 3.48 3.63 3.92 3.94 Nd ng/kg 7.61 9.11 12.2 12.5 14.8 14.8 Sm ng/kg 2.01 2.42 3.46 3.5 4.03 4.12 Eu ng/kg 0.5 0.51 0.77 0.77 0.95 0.93 Gd ng/kg 8.29 9.36 10.1 10.2 13.1 13 Tb ng/kg 0.37 0.37 0.57 0.59 0.7 0.72 Dy ng/kg 2.43 2.44 3.63 3.61 4.4 4.44 Ho ng/kg 0.56 0.57 0.79 0.79 0.95 0.98 Er ng/kg 1.8 1.91 2.34 2.44 2.86 2.91 Tm* ng/kg 0.27 0.27 0.34 0.35 0.42 0.43 Yb ng/kg 1.96 1.98 2.39 2.44 3.01 3.08 Lu ng/kg 0.35 0.37 0.4 0.41 0.51 0.53

Laanthr ng/kg 40.9 112 83.1 78.7 82.4 81.9

Lanat ng/kg 4.18 4.97 6.07 6.27 7.73 7.45

Gdanthr ng/kg 5.57 6.08 5.07 5.17 7.47 7.13

Gdnat ng/kg 2.72 3.29 5.03 5.04 5.63 5.91 ΣREE ng/kg 77.7 155 138 135 150 148 ΣREE (Natural) ng/kg 31.2 37.7 49.5 50.7 59.9 59.1 (La + Gd) / ΣREE % 60 76 64 62 60 60 * LaSN / La SN 10.8 23.5 14.7 13.5 11.7 12 * CeSN / Ce SN 0.41 0.48 0.48 0.49 0.48 0.44 * EuSN / Eu SN 1.01 0.84 0.87 0.87 0.94 0.89 * GdSN / Gd SN 3.04 2.85 2.01 2.03 2.33 2.21

NdSN/YbSN 0.32 0.38 0.43 0.43 0.41 0.4

TbSN/YbSN 0.69 0.68 0.88 0.87 0.84 0.85

177

Appendix 3 continued.

Sample Rhine/Wupper Rhine/Erft Rhine/Ruhr Rhine/Lippe Rhine Type Rhine Rhine Rhine Rhine Rhine Rhine‐km 703 735.4 779.4 811.4 857.3 Date 29/05/2008 29/05/2008 30/05/2008 30/05/2008 30/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 Weekday Thu Thu Fri Fri Fri Time 14:10 15:40 11:05 12:50 14:45 Latitude 51.04346 51.18311 51.44576 51.62743 51.83772 Longitude 6.94483 6.73364 6.7157 6.58483 6.17026 Temperature oC 21 21.1 21.1 21 20.9 pH 7.98 8.13 8.03 7.97 8.03 EC µS/cm 570 615 610 ‐ ‐ discharge m3/s 1800 1810 1860 1890 1940 Na+ mg/kg ‐ ‐‐‐‐ Mg2+ mg/kg ‐ ‐‐‐‐ Si mg/kg ‐ ‐‐‐‐ Cl‐ mg/kg ‐ ‐‐‐‐ K+ mg/kg ‐ ‐‐‐‐ Ca2+ mg/kg ‐ ‐‐‐‐ ‐ HCO3 mg/kg ‐ ‐‐‐‐ ‐ NO3 mg/kg ‐ ‐‐‐‐ Br‐ mg/kg ‐ ‐‐‐‐ 2‐ SO4 mg/kg ‐ ‐‐‐‐ Rb µg/kg 2.82 23.7 3.2 2.95 3.14 Sr mg/kg 0.36 0.33 0.36 0.37 0.38 Y ng/kg 26.8 26.8 32.7 34.1 36.6 Ba µg/kg 47.9 121 43.4 49.1 49.4 U µg/kg 0.84 0.29 0.76 0.82 0.85 La ng/kg 41 50.7 37.4 22.2 21.1 Ce ng/kg 6.23 5.83 5.51 5.08 4.7 Pr ng/kg 2.61 2.5 2.32 2.05 2 Nd ng/kg 10.3 9.74 9.36 8.23 8.56 Sm ng/kg 3.03 2.64 2.69 2.41 2.47 Eu ng/kg 0.69 0.61 0.66 0.59 0.6 Gd ng/kg 13.8 9.31 11.1 11.3 12.4 Tb ng/kg 0.52 0.47 0.49 0.47 0.46 Dy ng/kg 3.26 3.07 3.19 3.07 3.07 Ho ng/kg 0.75 0.69 0.73 0.74 0.7 Er ng/kg 2.33 2.12 2.26 2.42 2.26 Tm* ng/kg 0.36 0.33 0.34 0.37 0.35 Yb ng/kg 2.71 2.43 2.58 2.81 2.7 Lu ng/kg 0.51 0.44 0.46 0.51 0.5

Laanthr ng/kg 36.2 45.5 32.9 18.3 17

Lanat ng/kg 4.82 5.16 4.54 3.88 4.13

Gdanthr ng/kg 9.27 5.65 7.18 7.67 8.76

Gdnat ng/kg 4.57 3.66 3.96 3.61 3.64 ΣREE ng/kg 88.1 90.8 79.1 62.2 61.9 ΣREE (Natural) ng/kg 42.7 39.7 39.1 36.2 36.1 (La + Gd) / ΣREE % 52 56 51 42 42 * LaSN / La SN 8.5 9.81 8.24 5.71 5.12 * CeSN / Ce SN 0.46 0.42 0.44 0.47 0.41 * EuSN / Eu SN 0.87 0.92 0.95 0.94 0.94 * GdSN / Gd SN 3.03 2.54 2.81 3.12 3.41

NdSN/YbSN 0.32 0.33 0.3 0.24 0.26

TbSN/YbSN 0.70 0.70 0.69 0.61 0.63

178

Appendix 4 – May 2009: sample details and concentrations for the Rhine River

Sample Rhine Rhine/Thur Rhine/Töss Rhine/Wutach Rhine/Aare Rhine/Wiese Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 34.4 65.1 73.3 100.5 103 170.6 Date 03/05/2009 03/05/2009 03/05/2009 04/05/2009 04/05/2009 04/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Sun Sun Sun Mon Mon Mon Time 16:00 16:54 18:00 09:50 10:40 13:00 Latitude 47.69078 47.60646 47.55163 47.61611 47.60603 47.59248 Longitude 8.75336 8.59811 8.55629 8.25751 8.22718 7.59193 Temperature oC 13.6 12 12.4 12.5 12.6 13.8 pH 8.4 8.24 8.39 8.26 8.26 8.2 EC µS/cm 340 344 338 352 361 298 discharge m3/s 383 ‐ ‐ ‐ 969 1040 Na+ mg/kg 2.9 2.94 3.42 3.38 3.82 4.5 Mg2+ mg/kg 4.36 4.88 4.74 4.56 4.65 3.53 Si mg/kg 0.64 0.68 0.61 0.64 0.77 0.92 Cl‐ mg/kg 7.16 7.39 6.37 7.91 6.44 8 K+ mg/kg 0.72 0.74 0.83 0.83 0.86 0.87 Ca2+ mg/kg 25.5 26.2 25.8 27.2 27.5 24.1 ‐ HCO3 mg/kg ‐ NO3 mg/kg 4.02 4.56 4.29 4.29 4.64 4.62 Br‐ mg/kg 0.59 0.58 1.21 0.71 1.38 0.33 2‐ SO4 mg/kg 36.6 36.6 23.2 27.7 27.7 20.9 Rb µg/kg 1.39 1.22 1.48 1.53 1.36 2.03 Sr mg/kg 0.47 0.48 0.4 0.44 0.43 0.31 Y ng/kg 7.92 7.8 9.13 8.17 10.3 15.1 Ba µg/kg 27.4 30.3 28.9 30 32.2 37.3 U µg/kg 0.94 0.98 0.71 0.78 0.79 0.58 La ng/kg 1.08 1.08 1.37 1.37 1.82 5.61 Ce ng/kg 1.76 1.7 1.74 1.73 2.29 5.44 Pr ng/kg 0.28 0.26 0.37 0.36 0.47 1.5 Nd ng/kg 1.53 1.5 2.01 1.94 2.47 7.23 Sm ng/kg 0.69 0.69 0.75 0.71 0.89 1.78 Eu ng/kg 0.2 0.2 0.21 0.2 0.24 0.42 Gd ng/kg 1.85 1.89 1.91 1.81 2.11 3.19 Tb ng/kg 0.19 0.19 0.19 0.19 0.22 0.32 Dy ng/kg 1.21 1.24 1.27 1.23 1.46 2.08 Ho ng/kg 0.29 0.28 0.28 0.27 0.34 0.49 Er ng/kg 0.86 0.88 0.93 0.85 1.04 1.45 Tm* ng/kg 0.12 0.13 0.13 0.12 0.16 0.22 Yb ng/kg 0.88 0.92 0.93 0.89 1.14 1.66 Lu ng/kg 0.16 0.17 0.18 0.18 0.2 0.29

Laanthr ng/kg 0.7 0.72 0.71 0.71 0.97 1.2

Lanat ng/kg 0.38 0.36 0.67 0.66 0.86 4.42

Gdanthr ng/kg 0.28 0.26 0.49 0.49 0.48 0.95

Gdnat ng/kg 1.58 1.63 1.42 1.32 1.63 2.25 ΣREE ng/kg 11.1 11.1 12.3 11.8 14.8 31.7 ΣREE (Natural) ng/kg 10.1 10.1 11.1 10.6 13.4 29.6 (La + Gd) / ΣREE % 8.8 8.8 9.8 10 9.7 6.8 * LaSN / La SN 2.85 3 2.06 2.08 2.13 1.27 * CeSN / Ce SN 1.34 1.35 0.84 0.85 0.87 0.48 * EuSN / Eu SN 0.92 0.87 0.98 0.96 0.93 0.99 * GdSN / Gd SN 1.17 1.16 1.35 1.37 1.29 1.42

NdSN/YbSN 0.15 0.14 0.18 0.18 0.18 0.36

TbSN/YbSN 0.81 0.76 0.75 0.76 0.70 0.70

179

Appendix 4 continued.

Sample Rhine/Elz Rhine/Kinzig Rhine/Ill Rhine Rhine Rhine Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 260.9 294.2 304.6 354.3 372 414.3 Date 04/05/2009 04/05/2009 04/05/2009 05/05/2009 05/05/2009 05/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Mon Mon Mon Tue Tue Tue Time 14:05 16:15 17:20 11:10 13:05 16:00 Latitude 48.31209 48.57306 48.65078 48.97696 49.11572 49.43369 Longitude 7.71693 7.803647 7.86636 8.25663 8.36182 8.50634 Temperature oC 15.1 15.6 15.1 14.4 14.9 14.7 pH 8.27 8.17 8.2 7.96 8.05 8.02 EC µS/cm 360 375 360 377 392 390 discharge m3/s 1050 1030 1170 1180 1180 1150 Na+ mg/kg 5.57 6.02 6.28 6.61 6.68 6.89 Mg2+ mg/kg 4.14 4.03 3.94 3.96 4.06 4.11 Si mg/kg 0.7 0.77 0.74 0.82 0.81 0.82 Cl‐ mg/kg 13.6 15.4 13.6 19.3 15.7 14.1 K+ mg/kg 0.97 0.99 1.02 1.07 1.08 1.11 Ca2+ mg/kg 29 28.3 27.1 28.2 28.4 28.5 ‐ HCO3 mg/kg ‐ NO3 mg/kg 4.92 4.91 4.48 5.99 5.66 4.94 Br‐ mg/kg 0.47 0.46 0.61 0.5 0.38 0.63 2‐ SO4 mg/kg 25 24.7 23.9 28.8 23.5 23.2 Rb µg/kg 1.42 1.51 1.74 1.66 1.76 1.9 Sr mg/kg 0.38 0.38 0.36 0.36 0.38 0.38 Y ng/kg 10.4 10.8 12.5 16.2 16.3 15.3 Ba µg/kg 33.6 37.6 40.4 40.7 40.7 40.6 U µg/kg 0.66 0.74 0.65 0.68 0.77 0.73 La ng/kg 2.1 1.88 2.67 3.62 4.13 2.63 Ce ng/kg 2.14 2.09 2.65 4.62 4.33 3.57 Pr ng/kg 0.55 0.51 0.68 1.02 1 0.85 Nd ng/kg 2.74 2.67 3.56 5.23 5.14 4.07 Sm ng/kg 0.81 0.81 0.95 1.58 1.49 1.31 Eu ng/kg 0.22 0.22 0.24 0.37 0.36 0.32 Gd ng/kg 3.09 2.82 2.94 3.62 3.4 3.73 Tb ng/kg 0.17 0.18 0.21 0.32 0.3 0.26 Dy ng/kg 1.22 1.29 1.45 2.01 1.97 1.85 Ho ng/kg 0.28 0.31 0.34 0.47 0.45 0.44 Er ng/kg 0.91 0.98 1.11 1.49 1.48 1.37 Tm* ng/kg 0.14 0.16 0.17 0.22 0.22 0.22 Yb ng/kg 1.07 1.2 1.31 1.68 1.66 1.64 Lu ng/kg 0.19 0.23 0.23 0.3 0.31 0.29

Laanthr ng/kg 0.82 0.68 0.76 1.26 1.68 0.96

Lanat ng/kg 1.28 1.2 1.91 2.36 2.45 1.68

Gdanthr ng/kg 1.86 1.58 1.62 1.18 1.18 1.58

Gdnat ng/kg 1.23 1.25 1.31 2.44 2.22 2.14 ΣREE ng/kg 15.6 15.3 18.5 26.6 26.2 22.5 ΣREE (Natural) ng/kg 12.9 13.1 16.1 24.1 23.4 20 (La + Gd) / ΣREE % 17 15 13 9.2 11 11 * LaSN / La SN 1.64 1.56 1.4 1.54 1.69 1.57 * CeSN / Ce SN 0.6 0.62 0.52 0.69 0.64 0.73 * EuSN / Eu SN 1.02 1.02 1.03 0.89 0.92 0.91 * GdSN / Gd SN 2.52 2.27 2.24 1.48 1.53 1.74

NdSN/YbSN 0.21 0.19 0.23 0.26 0.26 0.21

TbSN/YbSN 0.59 0.55 0.59 0.69 0.65 0.57

180

Appendix 4 continued.

Sample Rhine/Pfrimm Rhine/Main Rhine/Nahe Rhine/Lahn Rhine/Ahr Rhine/Sieg Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 446.7 493 525.6 584.5 627.7 655.1 Date 05/05/2009 06/05/2009 06/05/2009 06/05/2009 06/05/2009 06/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Tue Wed Wed Wed Wed Wed Time 17:45 11:45 13:00 15:41 18:00 19:35 Latitude 49.66064 49.97127 49.96998 50.3004 50.54577 50.74 Longitude 8.36607 8.32867 7.93861 7.60238 7.28068 7.11326 Temperature oC 16.4 15.3 15.3 15.7 15.3 15.2 pH 7.88 7.92 8.03 8.01 8.07 8.06 EC µS/cm 514 480 492 505 625 503 discharge m3/s 1260 1260 1490 1460 1640 1630 Na+ mg/kg 12.3 9.64 10.2 11.2 15.6 11.1 Mg2+ mg/kg 4.29 5.15 5.08 5.92 7.15 5.87 Si mg/kg 0.86 0.88 0.83 0.95 0.79 0.94 Cl‐ mg/kg 38.3 26.2 137 37.2 44.4 31.2 K+ mg/kg 1.33 1.45 1.5 1.58 1.88 1.57 Ca2+ mg/kg 37.3 32.4 32.4 33.8 37.3 33.4 ‐ HCO3 mg/kg ‐ NO3 mg/kg 5.1 6.48 6.54 8 6.49 7.89 Br‐ mg/kg 0.4 0.43 0.47 0.5 0.34 0.41 2‐ SO4 mg/kg 43.1 38.8 42.6 46.7 38 38 Rb µg/kg 2.04 2.37 2.4 2.47 3.47 2.52 Sr mg/kg 0.38 0.4 0.41 0.43 0.46 0.41 Y ng/kg 15.6 17.3 18.5 20.1 21 18.7 Ba µg/kg 42.6 49.1 48 51.8 50.9 49.2 U µg/kg 0.71 0.78 0.83 0.75 0.82 0.74 La ng/kg 3.69 98.2 185 128 63.1 86 Ce ng/kg 3.46 8.16 7.2 6.31 5.83 5.92 Pr ng/kg 0.91 1.65 2.21 2.16 1.73 1.8 Nd ng/kg 4.48 6.75 7.86 8.35 7.14 7.27 Sm ng/kg 1.23 1.66 1.8 2.06 1.88 1.74 Eu ng/kg 0.33 0.39 0.44 0.5 0.47 0.46 Gd ng/kg 3.57 7.89 10.9 10.9 9.86 10.6 Tb ng/kg 0.26 0.33 0.33 0.38 0.39 0.35 Dy ng/kg 1.93 2.12 2.3 2.45 2.45 2.27 Ho ng/kg 0.49 0.49 0.52 0.59 0.62 0.53 Er ng/kg 1.6 1.59 1.73 1.84 2.35 1.8 Tm* ng/kg 0.25 0.24 0.25 0.28 0.38 0.28 Yb ng/kg 1.88 1.8 1.84 2.1 3.11 2.19 Lu ng/kg 0.34 0.37 0.33 0.42 0.67 0.46

Laanthr ng/kg 1.37 94 180 122 59.2 81.4

Lanat ng/kg 2.32 4.12 5.33 5.09 3.94 4.62

Gdanthr ng/kg 1.83 5.79 8.78 8.27 7.31 8.49

Gdnat ng/kg 1.74 2.1 2.12 2.6 2.55 2.14 ΣREE ng/kg 24.4 132 223 166 100 122 ΣREE (Natural) ng/kg 21.2 31.8 34.3 35.1 33.5 31.8 (La + Gd) / ΣREE % 13 76 85 79 66 74 * LaSN / La SN 1.59 23.8 34.7 25.1 16 18.6 * CeSN / Ce SN 0.55 0.77 0.55 0.48 0.56 0.51 * EuSN / Eu SN 1.05 0.97 1.07 1.02 1.01 1.11 * GdSN / Gd SN 2.05 3.75 5.13 4.18 3.87 4.96

NdSN/YbSN 0.2 0.31 0.36 0.33 0.19 0.28

TbSN/YbSN 0.50 0.67 0.66 0.67 0.45 0.58

181

Appendix 4 continued.

Sample Rhine/Wupper Rhine/Erft Rhine/Ruhr Rhine/Lippe Type Rhine Rhine Rhine Rhine Rhine‐km 703 735.4 779.4 811.4 Date 07/05/2009 07/05/2009 07/05/2009 07/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 Weekday Thu Thu Thu Thu Time 08:40 10:00 12:05 13:10 Latitude 51.04346 51.18311 51.44576 51.62743 Longitude 6.94483 6.73364 6.7157 6.58483 Temperature oC 15.2 15.8 15.9 15.5 pH 8.02 8.08 8.06 8.01 EC µS/cm 533 559 565 753 discharge m3/s 1670 1700 1750 1780 Na+ mg/kg 12.7 13.8 14 25.4 Mg2+ mg/kg 6.07 6.21 6.1 6.3 Si mg/kg 0.96 0.93 0.95 0.97 Cl‐ mg/kg 37.3 65.4 47.1 136 K+ mg/kg 1.73 1.75 1.84 1.91 Ca2+ mg/kg 33.9 36.6 34.8 41.1 ‐ HCO3 mg/kg ‐ NO3 mg/kg 7.67 8.19 8.83 8.79 Br‐ mg/kg 0.43 0.4 0.08 0.04 2‐ SO4 mg/kg 45.6 45.2 45.4 58.5 Rb µg/kg 2.75 3.34 3.38 3.87 Sr mg/kg 0.41 0.41 0.41 0.43 Y ng/kg 16.1 17.3 16.9 12.3 Ba µg/kg 49.3 49.2 49.4 56.4 U µg/kg 0.7 0.74 0.71 0.72 La ng/kg 80.6 68.7 54.9 21.5 Ce ng/kg 6.27 4.18 4.12 3.4 Pr ng/kg 1.86 1.48 1.44 0.85 Nd ng/kg 7.71 6.4 6.07 3.65 Sm ng/kg 2 1.63 1.53 1.04 Eu ng/kg 0.5 0.41 0.41 0.27 Gd ng/kg 15.3 11.3 13.8 12.3 Tb ng/kg 0.38 0.32 0.4 0.28 Dy ng/kg 2.51 2.19 2.8 2 Ho ng/kg 0.61 0.53 0.68 0.49 Er ng/kg 2.06 1.81 2.22 1.68 Tm* ng/kg 0.33 0.31 0.35 0.29 Yb ng/kg 2.57 2.55 2.65 2.37 Lu ng/kg 0.55 0.55 0.55 0.51

Laanthr ng/kg 76.3 64.9 51.4 19.8

Lanat ng/kg 4.37 3.71 3.58 1.78

Gdanthr ng/kg 12.6 9.21 11.8 10.7

Gdnat ng/kg 2.65 2.13 1.98 1.53 ΣREE ng/kg 123 102 91.9 50.6 ΣREE (Natural) ng/kg 34.4 28.2 28.8 20.1 (La + Gd) / ΣREE % 72 72 69 60 * LaSN / La SN 18.5 18.5 15.4 12.1 * CeSN / Ce SN 0.55 0.43 0.45 0.69 * EuSN / Eu SN 1.03 1.04 1.11 1.01 * GdSN / Gd SN 5.76 5.32 6.94 8.01

NdSN/YbSN 0.25 0.21 0.19 0.13

TbSN/YbSN 0.53 0.45 0.55 0.42

182

Appendix 4 continued.

Sample Rhine Rhine Rhine Type Rhine Rhine Rhine Rhine‐km 823.3 823.3 823.3 Date 07/05/2009 07/05/2009 07/05/2009 Sampling Campaign May 09 May 09 May 09 Weekday Thu Thu Thu Time 14:10 14:15 15:05 Latitude 51.6647 51.6647 51.6647 Longitude 6.48152 6.48152 6.48152 Temperature oC 16 16 16 pH 8.01 8.01 8.01 EC µS/cm 698 698 698 discharge m3/s 1840 1840 1840 Na+ mg/kg 22 22.3 21.6 Mg2+ mg/kg 6.27 6.31 6.26 Si mg/kg 0.98 0.98 0.97 Cl‐ mg/kg 68.4 67.1 67.5 K+ mg/kg 1.87 1.92 1.9 Ca2+ mg/kg 40.1 40.2 39.8 ‐ HCO3 mg/kg ‐ NO3 mg/kg 6.66 6.92 6.19 Br‐ mg/kg 0.29 0.22 0.29 2‐ SO4 mg/kg 59.1 57.2 59.5 Rb µg/kg 3.78 3.25 3.66 Sr mg/kg 0.43 0.43 0.42 Y ng/kg 12.7 14 13.2 Ba µg/kg 52.3 52.6 51.7 U µg/kg 0.73 0.72 0.71 La ng/kg 21.8 20.7 23.6 Ce ng/kg 3.86 3.34 3.75 Pr ng/kg 0.89 0.84 0.89 Nd ng/kg 3.97 3.71 4.18 Sm ng/kg 1.11 1.12 1.17 Eu ng/kg 0.29 0.29 0.28 Gd ng/kg 12.5 13.1 12.4 Tb ng/kg 0.25 0.25 0.25 Dy ng/kg 1.72 1.74 1.73 Ho ng/kg 0.45 0.44 0.43 Er ng/kg 1.54 1.53 1.51 Tm* ng/kg 0.28 0.28 0.28 Yb ng/kg 2.5 2.57 2.49 Lu ng/kg 0.46 0.5 0.56

Laanthr ng/kg 19.8 19 21.4

Lanat ng/kg 1.99 1.67 2.12

Gdanthr ng/kg 10.9 11.3 10.7

Gdnat ng/kg 1.6 1.73 1.67 ΣREE ng/kg 51.6 50.3 53.5 ΣREE (Natural) ng/kg 20.9 20 21.3 (La + Gd) / ΣREE % 59 60 60 * LaSN / La SN 10.9 12.4 11.1 * CeSN / Ce SN 0.71 0.71 0.65 * EuSN / Eu SN 1.02 0.98 0.96 * GdSN / Gd SN 7.8 7.55 7.43

NdSN/YbSN 0.13 0.12 0.14

TbSN/YbSN 0.36 0.35 0.37

183

Appendix 5 – October 2009: sample details and concentrations for the Rhine River

Sample Rhine Rhine/Thur Rhine/Töss Rhine/Wutach Rhine/Aare Rhine/Wiese Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 34.4 65.1 73.3 100.5 103 170.6 Date 25/10/2009 25/10/2009 25/10/2009 25/10/2009 25/10/2009 25/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Sun Sun Sun Sun Sun Sun Time 08:36 09:52 10:47 12:05 12:29 15:25 Latitude 47.69078 47.60646 47.55163 47.61611 47.60603 47.59248 Longitude 8.75336 8.59811 8.55629 8.25751 8.22718 7.59193 Temperature oC 12.5 13 12.6 12.8 12.8 13.4 pH 7.87 7.87 8.01 8.04 8.13 8.07 EC µS/cm 317 323 336 354 345 335 discharge m3/s 247 ‐ ‐ ‐ 598 ‐ Na+ mg/kg 3.74 3.84 4.02 4.89 4.97 7.14 Mg2+ mg/kg 7.67 8.53 7.96 9.04 8.69 7.09 Si mg/kg 0.61 0.67 0.62 0.73 0.69 0.73 Cl‐ mg/kg 12.2 13.6 14.4 17.7 18.3 24.5 K+ mg/kg 1.85 1.9 1.99 2.38 2.3 2.62 Ca2+ mg/kg 38.4 39.8 39.6 44.2 41.8 41.8 ‐ HCO3 mg/kg 68.5 66.2 72.3 72.8 71.5 66.8 ‐ NO3 mg/kg 4.85 6.88 6.66 8.33 8.11 8.93 Br‐ mg/kg 1.08 1.09 1.16 1.16 1.19 0.63 2‐ SO4 mg/kg 60.8 60.4 52 56.5 59.2 48 Rb µg/kg 1.1 1.16 1.2 1.25 1.34 1.75 Sr mg/kg 0.44 0.45 0.41 0.42 0.42 0.23 Y ng/kg 4.35 4.63 6.47 6.08 7.6 9.16 Ba µg/kg 26.1 31 29.4 31.9 31.4 30.9 U µg/kg 0.98 1.06 0.88 0.86 0.94 0.67 La ng/kg 0.97 0.84 1.18 1.11 1.6 2.71 Ce ng/kg 1.14 0.98 1.48 1.17 1.6 2.92 Pr ng/kg 0.2 0.17 0.33 0.3 0.42 0.71 Nd ng/kg 1.07 0.98 1.59 1.59 2.11 3.26 Sm ng/kg 0.44 0.4 0.57 0.55 0.72 0.87 Eu ng/kg 0.13 0.14 0.16 0.15 0.19 0.2 Gd ng/kg 36.5 51.2 39.1 17.5 20.2 18 Tb ng/kg 0.19 0.18 0.13 0.15 0.17 0.2 Dy ng/kg 0.78 0.78 0.86 0.9 1.12 1.12 Ho ng/kg 0.18 0.19 0.2 0.19 0.25 0.26 Er ng/kg 0.52 0.52 0.65 0.65 0.8 0.87 Tm* ng/kg 0.07 0.08 0.1 0.1 0.12 0.14 Yb ng/kg 0.54 0.63 0.77 0.73 0.92 1.1 Lu ng/kg 0.16 0.16 0.18 0.13 0.2 0.21

Laanthr ng/kg 0.67 0.56 0.62 0.53 0.81 0.96

Lanat ng/kg 0.3 0.28 0.55 0.58 0.79 1.75

Gdanthr ng/kg 35.5 50.4 38 16.5 18.9 16.8

Gdnat ng/kg 0.95 0.83 1.05 0.98 1.26 1.2 ΣREE ng/kg 42.9 57.3 47.3 25.2 30.4 32.6 ΣREE (Natural) ng/kg 6.67 6.31 8.64 8.17 10.7 14.8 (La + Gd) / ΣREE % 84 89 82 68 65 55 * LaSN / La SN 3.25 2.99 2.13 1.9 2.02 1.55 * CeSN / Ce SN 1.15 1.06 0.87 0.66 0.67 0.62 * EuSN / Eu SN 0.92 1.11 1 0.95 0.92 0.94 * GdSN / Gd SN 38.6 61.8 37.1 17.9 16 15

NdSN/YbSN 0.17 0.13 0.17 0.18 0.19 0.25

TbSN/YbSN 1.28 1.05 0.63 0.77 0.66 0.68

184

Appendix 5 continued.

Sample Rhine/Elz Rhine/Kinzig Rhine/Ill Rhine Rhine Rhine/Pfrimm Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 260.9 294.2 304.6 354.3 414.3 446.7 Date 25/10/2009 26/10/2009 26/10/2009 26/10/2009 26/10/2009 27/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Sun Mon Mon Mon Mon Tue Time 16:40 09:29 10:42 14:12 16:38 09:30 Latitude 48.31209 48.57306 48.65078 48.97696 49.43369 49.66064 Longitude 7.71693 7.803647 7.86636 8.25663 8.50634 8.36607 Temperature oC 13.2 12.7 13.3 13.1 14.3 15.4 pH 7.98 7.92 7.78 7.94 7.96 7.69 EC µS/cm 385 390 381 450 419 644 discharge m3/s 932 805 814 656 690 727 Na+ mg/kg 10.4 10.8 11.1 11.7 12.4 29.3 Mg2+ mg/kg 7.4 7.36 7.31 7.55 7.74 8.76 Si mg/kg 0.71 0.83 0.82 0.9 0.96 1.17 Cl‐ mg/kg 45.1 48.2 48.3 27.4 27.2 85.7 K+ mg/kg 3.06 3.04 3.27 3.41 3.62 5.07 Ca2+ mg/kg 43.9 44.4 43.7 45.1 45.7 62.9 ‐ HCO3 mg/kg 66.6 66 61.5 107 115 73.1 ‐ NO3 mg/kg 10.4 10.3 10.7 5.55 5.48 6.23 Br‐ mg/kg 0.67 0.67 0.66 0.2 0.19 0.23 2‐ SO4 mg/kg 53 53.2 52.7 27.2 27.2 63.1 Rb µg/kg 1.78 1.62 1.86 1.97 2.08 2.3 Sr mg/kg 0.37 0.37 0.34 0.36 0.36 0.39 Y ng/kg 9.97 11.3 12 13 15 14.2 Ba µg/kg 38.1 40.8 41.5 41.8 43.8 53.5 U µg/kg 0.76 0.78 0.7 0.73 0.75 0.82 La ng/kg 2.82 2.85 3.69 2.03 1.91 2.43 Ce ng/kg 3.71 3.43 4.42 2.13 2.09 2.15 Pr ng/kg 0.63 0.73 0.84 0.59 0.75 0.73 Nd ng/kg 3.09 3.6 3.89 3.06 3.93 3.92 Sm ng/kg 0.76 1.1 1.12 0.96 1.14 1.21 Eu ng/kg 0.22 0.26 0.29 0.27 0.32 0.33 Gd ng/kg 11.9 9.12 10.2 10.9 11.8 13.1 Tb ng/kg 0.19 0.24 0.23 0.35 0.29 0.3 Dy ng/kg 1.17 1.39 1.47 2.69 1.9 1.86 Ho ng/kg 0.29 0.3 0.34 0.74 0.46 0.49 Er ng/kg 0.9 1.08 1.07 2.74 1.51 1.59 Tm* ng/kg 0.14 0.17 0.17 0.43 0.25 0.26 Yb ng/kg 1.06 1.34 1.27 3.32 1.98 2.09 Lu ng/kg 0.18 0.31 0.23 0.73 0.37 0.35

Laanthr ng/kg 0.93 1.25 1.8 0.73 0.04 0.72

Lanat ng/kg 1.88 1.6 1.89 1.3 1.87 1.7

Gdanthr ng/kg 10.9 7.41 8.61 9.38 10 11.2

Gdnat ng/kg 0.96 1.71 1.64 1.54 1.71 1.92 ΣREE ng/kg 27 25.9 29.3 30.9 28.7 30.8 ΣREE (Natural) ng/kg 15.2 17.3 18.9 20.8 18.6 18.9 (La + Gd) / ΣREE % 44 33 36 33 35 39 * LaSN / La SN 1.49 1.78 1.95 1.56 1.02 1.42 * CeSN / Ce SN 0.77 0.76 0.85 0.57 0.4 0.44 * EuSN / Eu SN 1.21 0.89 1 1.04 1.06 1.03 * GdSN / Gd SN 12.4 5.33 6.24 7.1 6.88 6.84

NdSN/YbSN 0.24 0.22 0.25 0.08 0.17 0.16

TbSN/YbSN 0.67 0.65 0.67 0.38 0.54 0.52

185

Appendix 5 continued.

Sample Rhine/Main Rhine/Nahe Rhine/Lahn Rhine/Mosel Rhine/Ahr Rhine/Sieg Type Rhine Rhine Rhine Rhine Rhine Rhine Rhine‐km 493 525.6 584.5 591.5 627.7 655.1 Date 27/10/2009 27/10/2009 27/10/2009 27/10/2009 28/10/2009 28/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 ‐ Oct 09 Weekday Tue Tue Tue Tue ‐ Wed Time 12:52 14:02 16:54 18:01 09:58 11:36 Latitude 49.97127 49.96998 50.3004 50.35874 50.54577 50.74 Longitude 8.32867 7.93861 7.60238 7.60476 7.28068 7.11326 Temperature oC 13.6 14.1 13.3 12.7 12.3 12.8 pH 7.85 8.02 7.96 7.81 7.87 8.13 EC µS/cm 540 549 594 575 738 589 discharge m3/s 727 886 883 894 ‐ 962 Na+ mg/kg 18.8 19.6 22.1 20.9 33.9 21.8 Mg2+ mg/kg 9.35 9.45 11.7 11.8 19.4 11.8 Si mg/kg 1.11 0.97 1.22 1.51 1.05 1.21 Cl‐ mg/kg 48.1 53.9 55.3 58.7 72 35.6 K+ mg/kg 5.23 5.1 5.76 5.86 8.68 5.7 Ca2+ mg/kg 59 61.4 65.3 61 54.7 64.3 ‐ HCO3 mg/kg 116 118 112 114 146 183 ‐ NO3 mg/kg 7.38 6.91 8.53 8.42 19.1 2.77 Br‐ mg/kg 0.2 0.2 0.17 0.16 0.67 0.29 2‐ SO4 mg/kg 49.4 50.7 58.8 52.4 34.1 39.3 Rb µg/kg 2.75 2.6 3.25 3.37 7.2 4.27 Sr mg/kg 0.41 0.41 0.44 0.4 0.16 0.1 Y ng/kg 19.1 19.9 17.7 17.3 10 21.5 Ba µg/kg 47.6 46.3 53.1 47.9 16.8 18.3 U µg/kg 0.74 0.82 0.87 0.79 0.29 0.06 La ng/kg 338 294 195 209 11 171 Ce ng/kg 8.39 8.09 8.32 6.1 1.54 10.1 Pr ng/kg 2.65 2.72 2.05 1.97 0.4 2.47 Nd ng/kg 9.15 8.64 7.46 7.16 2.07 9.08 Sm ng/kg 1.88 1.86 1.73 1.67 0.68 2.16 Eu ng/kg 0.48 0.48 0.44 0.46 0.19 0.57 Gd ng/kg 27.2 20.9 29.7 32.5 126 28.2 Tb ng/kg 0.39 0.38 0.36 0.35 0.17 0.44 Dy ng/kg 2.43 2.44 2.2 2.1 1.23 2.67 Ho ng/kg 0.57 0.56 0.52 0.5 0.38 0.63 Er ng/kg 1.9 1.86 1.78 1.64 1.67 1.97 Tm* ng/kg 0.29 0.29 0.28 0.26 0.32 0.31 Yb ng/kg 2.3 2.19 2.21 2.1 3.19 2.37 Lu ng/kg 0.47 0.53 0.44 0.45 0.68 0.42

Laanthr ng/kg 331 288 190 205 10.2 165

Lanat ng/kg 7.3 6.45 4.96 4.73 0.82 5.83

Gdanthr ng/kg 25.2 18.9 27.6 30.5 125 25.5

Gdnat ng/kg 1.99 2.06 2.07 2 1.16 2.65 ΣREE ng/kg 396 345 253 267 150 232 ΣREE (Natural) ng/kg 40.2 38.5 34.8 31.5 14.5 41.6 (La + Gd) / ΣREE % 90 89 86 88 82 * LaSN / La SN 46.3 45.6 39.4 44.3 13.6 29.4 * CeSN / Ce SN 0.49 0.52 0.68 0.52 0.64 0.69 * EuSN / Eu SN 1.17 1.16 1.1 1.17 0.99 1.11 * GdSN / Gd SN 13.7 10.2 14.3 16.2 109 10.7

NdSN/YbSN 0.33 0.33 0.28 0.28 0.05 0.32

TbSN/YbSN 0.62 0.63 0.59 0.60 0.19 0.68

186

Appendix 5 continued.

Sample Rhine/Wupper Rhine/Erft Rhine/Ruhr Rhine/Lippe Type Rhine Rhine Rhine Rhine Rhine‐km 703 735.4 779.4 811.4 Date 28/10/2009 28/10/2009 28/10/2009 28/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Weekday Wed Wed Wed Wed Time 14:19 15:45 17:10 18:47 Latitude 51.04346 51.18311 51.44576 51.62743 Longitude 6.94483 6.73364 6.7157 6.58483 Temperature oC 13.8 13.4 14.1 13.3 pH 7.94 7.94 7.9 7.86 EC µS/cm 647 701 680 1324 discharge m3/s 986 1010 1060 1050 Na+ mg/kg 25.1 27.6 21.6 38.3 Mg2+ mg/kg 11.5 12.6 12.4 12.3 Si mg/kg 1.27 1.29 1.24 1.29 Cl‐ mg/kg 43.2 53.9 52.9 96.1 K+ mg/kg 6.34 6.42 6.33 6.33 Ca2+ mg/kg 62.4 70.7 69.9 75.8 ‐ HCO3 mg/kg 178 174 187 179 ‐ NO3 mg/kg 4.05 3.57 3.72 5.93 Br‐ mg/kg 0.28 0.3 0.32 0.33 2‐ SO4 mg/kg 44.4 46.8 49.2 59.9 Rb µg/kg 3.49 3.67 4.15 4.26 Sr mg/kg 0.44 0.47 0.45 0.48 Y ng/kg 13.3 15.6 16.1 17.8 Ba µg/kg 48.6 51.4 50.7 54.6 U µg/kg 0.78 0.8 0.73 0.79 La ng/kg 71.7 80 69.6 28.6 Ce ng/kg 5.08 4.2 4.19 3.53 Pr ng/kg 1.29 1.34 1.35 0.94 Nd ng/kg 5.02 5.26 5.67 4.19 Sm ng/kg 1.29 1.43 1.56 1.15 Eu ng/kg 0.35 0.37 0.41 0.32 Gd ng/kg 36.1 29.4 28.2 27 Tb ng/kg 0.28 0.29 0.3 0.27 Dy ng/kg 1.67 1.9 1.94 1.8 Ho ng/kg 0.41 0.47 0.46 0.47 Er ng/kg 1.52 1.7 1.62 1.73 Tm* ng/kg 0.27 0.28 0.27 0.29 Yb ng/kg 2.44 2.31 2.3 2.53 Lu ng/kg 0.49 0.44 0.66 0.61

Laanthr ng/kg 68.8 77.2 66.7 26.4

Lanat ng/kg 2.86 2.77 2.93 2.17

Gdanthr ng/kg 34.3 27.4 26 25.4

Gdnat ng/kg 1.71 2 2.2 1.62 ΣREE ng/kg 128 129 119 73.5 ΣREE (Natural) ng/kg 24.7 24.8 25.9 21.6 (La + Gd) / ΣREE % 81 81 78 71 * LaSN / La SN 25.1 28.9 23.8 13.2 * CeSN / Ce SN 0.68 0.57 0.53 0.6 * EuSN / Eu SN 1.1 1.02 1.04 1.09 * GdSN / Gd SN 21.1 14.7 12.8 16.7

NdSN/YbSN 0.17 0.19 0.2 0.14

TbSN/YbSN 0.41 0.46 0.48 0.38

187

Appendix 5 continued.

Sample Rhine Rhine Rhine Type Rhine Rhine Rhine Rhine‐km 823.3 823.3 823.3 Date 28/10/2009 28/10/2009 28/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Weekday Wed Wed Wed Time 20:00 20:00 20:00 Latitude 51.6647 51.6647 51.6647 Longitude 6.48152 6.48152 6.48152 Temperature oC 12.5 12.5 12.5 pH 7.96 7.96 7.96 EC µS/cm 725 725 725 discharge m3/s 1120 1120 1120 Na+ mg/kg 34.6 34.7 33.4 Mg2+ mg/kg 12.7 12.8 12.6 Si mg/kg 1.32 1.29 1.27 Cl‐ mg/kg 104 102 104 K+ mg/kg 6.5 6.63 6.33 Ca2+ mg/kg 76.5 75.8 77.8 ‐ HCO3 mg/kg 175 178 177 ‐ NO3 mg/kg 4.41 4.57 4.74 Br‐ mg/kg 0.33 0.33 0.33 2‐ SO4 mg/kg 55.6 56.5 57.9 Rb µg/kg 4.29 4.72 4.3 Sr mg/kg 0.47 0.49 0.47 Y ng/kg 19.5 19.9 19.8 Ba µg/kg 55.8 55.2 54.7 U µg/kg 0.81 0.77 0.78 La ng/kg 38.9 43.2 41.3 Ce ng/kg 4.47 4.91 4.34 Pr ng/kg 1.01 1.11 1.14 Nd ng/kg 4.53 4.98 5.08 Sm ng/kg 1.28 1.31 1.33 Eu ng/kg 0.33 0.36 0.33 Gd ng/kg 27.9 25.4 26.9 Tb ng/kg 0.31 0.29 0.29 Dy ng/kg 2 1.91 1.97 Ho ng/kg 0.5 0.52 0.54 Er ng/kg 1.89 1.86 1.83 Tm* ng/kg 0.31 0.3 0.31 Yb ng/kg 2.65 2.5 2.54 Lu ng/kg 0.6 0.54 0.53

Laanthr ng/kg 36.7 40.5 38.5

Lanat ng/kg 2.24 2.77 2.84

Gdanthr ng/kg 26 23.6 25.1

Gdnat ng/kg 1.86 1.76 1.78 ΣREE ng/kg 86.7 89.2 88.5 ΣREE (Natural) ng/kg 24 25.1 24.9 (La + Gd) / ΣREE % 72 72 72 * LaSN / La SN 17.4 15.6 14.5 * CeSN / Ce SN 0.73 0.67 0.58 * EuSN / Eu SN 1.01 1.11 1 * GdSN / Gd SN 15 14.4 15.1

NdSN/YbSN 0.14 0.17 0.17

TbSN/YbSN 0.43 0.42 0.42

188 Appendix 6 – May 2008: sample details and concentrations for the tributaries to the Rhine River

Sample Thur Töss Wutach Aare Birs Wiese Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 66.6 73.4 101 103.2 164 169.2 Date 25/05/2008 25/05/2008 25/05/2008 25/05/2008 26/05/2008 26/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Sun Sun Sun Sun Mon Mon Time 12:40 13:25 14:55 16:00 09:45 12:40 Latitude 47.59429 47.54998 47.61944 47.60036 47.5495 47.58187 Longitude 8.60689 8.55497 8.25878 8.22022 7.62251 7.59182 Temperature oC 17.1 17.3 15 16.4 14.4 15.5 pH 8.38 8.32 8.38 8.2 8.19 7.85 EC µS/cm 461 670 615 346 449 157 discharge m3/s 23.8 2.9 5.84 608 8.07 5.28 Na+ mg/kg ‐‐‐‐‐‐ Mg2+ mg/kg ‐‐‐‐‐‐ Si mg/kg ‐‐‐‐‐‐ Cl‐ mg/kg ‐‐‐‐‐‐ K+ mg/kg ‐‐‐‐‐‐ Ca2+ mg/kg ‐‐‐‐‐‐ ‐ HCO3 mg/kg ‐‐‐‐‐‐ ‐ NO3 mg/kg ‐‐‐‐‐‐ Br‐ mg/kg ‐‐‐‐‐‐ 2‐ SO4 mg/kg ‐‐‐‐‐‐ Rb µg/kg 1.27 2.36 2.47 1.18 1.47 3.41 Sr mg/kg 0.2 0.24 0.42 0.29 0.19 0.07 Y ng/kg 8.88 4 19.4 7 10.4 21.2 Ba µg/kg 42.8 58.5 87.9 31.9 22.9 64.3 U µg/kg 0.79 1.37 0.75 0.56 0.41 0.22 La ng/kg 2.23 0.89 3.75 1.69 3.16 12.5 Ce ng/kg 3.45 1.46 4.88 2.12 3.56 11.5 Pr ng/kg 0.6 0.21 1.55 0.39 0.79 3.2 Nd ng/kg 2.65 0.91 7.61 1.83 3.46 14.5 Sm ng/kg 0.6 0.24 2.06 0.43 0.75 3.29 Eu ng/kg 0.16 0.07 0.52 0.11 0.19 0.71 Gd ng/kg 3.04 4.28 4.54 1.34 1.25 4.21 Tb ng/kg 0.14 0.06 0.38 0.09 0.14 0.47 Dy ng/kg 0.95 0.44 2.38 0.72 1.01 2.95 Ho ng/kg 0.24 0.11 0.52 0.18 0.23 0.71 Er ng/kg 0.83 0.4 1.65 0.61 0.76 2.62 Tm* ng/kg 0.13 0.07 0.22 0.09 0.11 0.46 Yb ng/kg 1.07 0.55 1.51 0.66 0.83 4.06 Lu ng/kg 0.21 0.1 0.27 0.12 0.15 0.88

Laanthr ng/kg 0.4 0.39 ‐0.3 0.49 0.62 2.5

Lanat ng/kg 1.84 0.5 4.02 1.2 2.54 10

Gdanthr ng/kg 2.35 3.95 1.67 0.82 0.4 0.38

Gdnat ng/kg 0.7 0.33 2.87 0.52 0.84 3.83 ΣREE ng/kg 16.3 9.79 31.8 10.4 16.4 62.1 ΣREE (Natural) ng/kg 13.6 5.45 30.4 9.07 15.4 59.2 (La + Gd) / ΣREE % 17 44 4.4 13 6.2 4.6 * LaSN / La SN 1.22 1.78 0.93 1.41 1.24 1.25 * CeSN / Ce SN 0.77 1.1 0.45 0.71 0.58 0.47 * EuSN / Eu SN 1.13 1.15 1.01 1.11 1.11 0.95 * GdSN / Gd SN 4.37 13.1 1.58 2.59 1.48 1.1

NdSN/YbSN 0.21 0.14 0.42 0.23 0.35 0.3

TbSN/YbSN 0.49 0.43 0.92 0.52 0.62 0.42

189

Appendix 6 continued.

Sample Elz Kinzig Ill Acher/Rench Moder Murg Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 267.5 299.4 312 315.3 334.4 344.6 Date 26/05/2008 26/05/2008 27/05/2008 27/05/2008 27/05/2008 27/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Mon Mon Tue Tue Tue Tue Time 15:45 18:50 10:00 10:40 12:20 14:05 Latitude 48.29868 48.57573 48.68879 48.70663 48.80944 48.9035 Longitude 7.74147 7.82399 7.90925 7.96121 8.05641 8.18033 Temperature oC 18.2 19 18.5 19.6 18.9 18.5 pH 7.93 7.57 7.92 7.41 7.72 8.23 EC µS/cm 219 218 466 273 467 168 discharge m3/s ‐ 8.62 ‐ ‐ ‐ 6.18 Na+ mg/kg ‐ ‐‐‐‐ ‐ Mg2+ mg/kg ‐ ‐‐‐‐ ‐ Si mg/kg ‐ ‐‐‐‐ ‐ Cl‐ mg/kg ‐ ‐‐‐‐ ‐ K+ mg/kg ‐ ‐‐‐‐ ‐ Ca2+ mg/kg ‐ ‐‐‐‐ ‐ ‐ HCO3 mg/kg ‐ ‐‐‐‐ ‐ ‐ NO3 mg/kg ‐ ‐‐‐‐ ‐ Br‐ mg/kg ‐ ‐‐‐‐ ‐ 2‐ SO4 mg/kg ‐ ‐‐‐‐ ‐ Rb µg/kg 2.51 5.92 2.02 7.11 4.02 5.64 Sr mg/kg 0.09 0.07 0.29 0.13 0.35 0.05 Y ng/kg 12.9 15.9 21 15.7 20.7 67.4 Ba µg/kg 68.4 125 74 85.3 71.9 109 U µg/kg 0.23 0.22 0.87 0.27 0.68 0.11 La ng/kg 6.57 5.02 4.18 14.1 2.88 9.05 Ce ng/kg 7.45 5.56 5 18.9 3.78 12.3 Pr ng/kg 1.62 1.46 1.18 2.25 0.85 4.16 Nd ng/kg 7.21 6.8 5.8 7.47 3.99 21 Sm ng/kg 1.79 1.92 1.62 1.51 1.2 6.93 Eu ng/kg 0.41 0.5 0.38 0.4 0.37 1.72 Gd ng/kg 3.69 7.71 4.09 5.6 5.48 17.9 Tb ng/kg 0.29 0.35 0.34 0.27 0.33 1.41 Dy ng/kg 1.98 2.27 2.59 1.81 2.21 8.29 Ho ng/kg 0.45 0.53 0.68 0.44 0.7 1.79 Er ng/kg 1.41 1.73 2.48 1.6 2.57 5.52 Tm* ng/kg 0.24 0.27 0.41 0.27 0.45 0.78 Yb ng/kg 1.98 2.07 3.38 2.33 3.88 5.49 Lu ng/kg 0.34 0.38 0.67 0.45 0.82 0.83

Laanthr ng/kg 2.22 1.65 1.24 7.97 1.07 0.76

Lanat ng/kg 4.35 3.37 2.93 6.14 1.81 8.3

Gdanthr ng/kg 1.41 4.92 1.76 4.04 3.63 6.18

Gdnat ng/kg 2.28 2.79 2.32 1.56 1.84 11.7 ΣREE ng/kg 35.5 36.6 32.8 57.4 29.5 97.2 ΣREE (Natural) ng/kg 31.8 30 29.8 45.4 24.8 90.2 (La + Gd) / ΣREE % 10 18 9.2 21 16 7.1 * LaSN / La SN ‐ 1.51 1.49 1.42 2.30 1.59 1.09 * CeSN / Ce SN ‐ 0.67 0.60 0.63 1.33 0.74 0.50 * EuSN / Eu SN ‐ 0.96 1.01 0.93 1.23 1.18 0.90 * GdSN / Gd SN ‐ 1.62 2.76 1.76 3.58 2.97 1.53

NdSN/YbSN ‐ 0.30 0.27 0.14 0.27 0.09 0.32

TbSN/YbSN ‐ 0.54 0.61 0.37 0.43 0.31 0.94

190

Appendix 6 continued.

Sample Neckar Main Nahe Lahn Mosel Wied Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 428.2 496.7 529.2 585.8 592.3 610.3 Date 28/05/2008 28/05/2008 28/05/2008 28/05/2008 28/05/2008 29/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Wed Wed Wed Wed Wed Thu Time 10:50 12:50 14:30 16:55 16:15 09:35 Latitude 49.49463 50.00068 49.9669 50.30885 50.3631 50.44075 Longitude 8.46724 8.41915 7.89106 7.60545 7.60081 7.45165 Temperature oC 19.9 20.9 21.8 20.2 20.7 19.5 pH 8 8.46 7.97 8.82 8.21 7.77 EC µS/cm 748 776 580 495 1148 315 discharge m3/s 85.4 153 8.35 17.4 145 2.43 Na+ mg/kg ‐‐‐‐‐ ‐ Mg2+ mg/kg ‐‐‐‐‐ ‐ Si mg/kg ‐‐‐‐‐ ‐ Cl‐ mg/kg ‐‐‐‐‐ ‐ K+ mg/kg ‐‐‐‐‐ ‐ Ca2+ mg/kg ‐‐‐‐‐ ‐ ‐ HCO3 mg/kg ‐‐‐‐‐ ‐ ‐ NO3 mg/kg ‐‐‐‐‐ ‐ Br‐ mg/kg ‐‐‐‐‐ ‐ 2‐ SO4 mg/kg ‐‐‐‐‐ ‐ Rb µg/kg 4.11 4.42 5.33 4.33 4.64 3.64 Sr mg/kg 0.63 0.54 0.25 0.16 0.5 0.1 Y ng/kg 19.4 127 23.8 117 18.6 5.85 Ba µg/kg 112 105 130 31.3 98.5 24.2 U µg/kg 0.82 1 1.23 0.37 0.56 0.09 La ng/kg 2.11 18.8 8.4 3.16 9.25 1.94 Ce ng/kg 2.52 39.5 5.7 2.04 10.1 2.09 Pr ng/kg 0.7 11.8 1.72 0.52 1.34 0.51 Nd ng/kg 3.55 56.4 8.43 2.46 6.39 2.14 Sm ng/kg 0.93 15.2 2.21 0.61 1.58 0.63 Eu ng/kg 0.28 3.84 0.56 0.19 0.42 0.19 Gd ng/kg 24.9 43.2 35.7 21.5 7.94 9.6 Tb ng/kg 0.24 2.62 0.48 0.15 0.32 0.11 Dy ng/kg 1.64 15.3 3.68 1 2.19 0.67 Ho ng/kg 0.44 3.26 0.95 0.28 0.58 0.18 Er ng/kg 1.48 9.07 3.36 0.98 2.15 0.54 Tm* ng/kg 0.22 1.16 0.5 0.17 0.33 0.1 Yb ng/kg 1.59 7.4 3.77 1.48 2.66 0.86 Lu ng/kg 0.26 1.12 0.67 0.3 0.48 0.16

Laanthr ng/kg 0.13 ‐11 3.71 1.66 5.39 0.96

Lanat ng/kg 1.98 30.1 4.68 1.5 3.86 0.98

Gdanthr ng/kg 23.7 22.2 32.7 20.8 5.93 8.63

Gdnat ng/kg 1.24 21 2.98 0.76 2.02 0.97 ΣREE ng/kg 40.9 229 76.2 34.9 45.7 19.7 ΣREE (Natural) ng/kg 17.1 218 39.7 12.4 34.4 10.1 (La + Gd) / ΣREE % 58 4.8 48 64 25 49 * LaSN / La SN ‐ 1.06 0.63 1.79 2.11 2.40 1.98 * CeSN / Ce SN ‐ 0.48 0.49 0.46 0.53 1.02 0.76 * EuSN / Eu SN ‐ 1.21 1.01 1.03 1.29 1.10 1.12 * GdSN / Gd SN ‐ 20.10 2.06 12.00 28.20 3.94 9.92

NdSN/YbSN ‐ 0.19 0.63 0.19 0.14 0.20 0.21

TbSN/YbSN ‐ 0.54 1.29 0.47 0.38 0.44 0.47

191

Appendix 6 continued.

Sample Ahr Sieg Wupper Erft Ruhr Ruhr Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 629.3 659.5 703.4 735.6 780.7 780.7 Date 29/05/2008 29/05/2008 29/05/2008 29/05/2008 30/05/2008 30/05/2008 Sampling Campaign May 08 May 08 May 08 May 08 May 08 May 08 Weekday Thu Thu Thu Thu Fri Fri Time 10:45 12:20 14:20 15:15 11:20 11:40 Latitude 50.54976 50.76556 51.04565 51.18269 51.44582 51.44405 Longitude 7.24459 7.10716 6.94844 6.73103 6.72884 6.73561 Temperature oC 18.2 17.6 18.9 23.1 19.8 19.8 pH 7.58 7.47 7.81 7.94 7.57 7.6 EC µS/cm 498 306 435 793 593 594 discharge m3/s 1.53 20.2 6.64 9.8 23.8 23.8 Na+ mg/kg ‐ ‐‐‐‐ ‐ Mg2+ mg/kg ‐ ‐‐‐‐ ‐ Si mg/kg ‐ ‐‐‐‐ ‐ Cl‐ mg/kg ‐ ‐‐‐‐ ‐ K+ mg/kg ‐ ‐‐‐‐ ‐ Ca2+ mg/kg ‐ ‐‐‐‐ ‐ ‐ HCO3 mg/kg ‐ ‐‐‐‐ ‐ ‐ NO3 mg/kg ‐ ‐‐‐‐ ‐ Br‐ mg/kg ‐ ‐‐‐‐ ‐ 2‐ SO4 mg/kg ‐ ‐‐‐‐ ‐ Rb µg/kg 2.98 3.46 3.92 26 4.16 4.22 Sr mg/kg 0.15 0.09 0.09 0.35 0.23 0.24 Y ng/kg 20.6 9.48 13.9 10.5 22.6 25.4 Ba µg/kg 21.8 17.1 21.1 141 39.8 39.8 U µg/kg 0.27 0.04 0.1 0.32 0.18 0.17 La ng/kg 5.06 2.87 3.29 2.73 2.09 1.96 Ce ng/kg 2.49 2.12 3.02 1.88 2.57 2.7 Pr ng/kg 1.12 0.58 0.8 0.44 0.97 1.09 Nd ng/kg 5.91 2.8 3.58 2.02 6.11 6.69 Sm ng/kg 1.84 0.75 1.06 0.52 2.01 2.25 Eu ng/kg 0.51 0.24 0.27 0.15 0.55 0.63 Gd ng/kg 4.96 25.5 60.4 14.2 13.8 15.8 Tb ng/kg 0.35 0.16 0.19 0.14 0.46 0.48 Dy ng/kg 2.18 1.04 1.19 0.93 2.96 3.14 Ho ng/kg 0.53 0.28 0.33 0.23 0.74 0.77 Er ng/kg 1.87 1.09 1.27 0.8 2.74 2.72 Tm* ng/kg 0.31 0.2 0.24 0.13 0.46 0.43 Yb ng/kg 2.62 1.93 2.27 1.02 3.87 3.46 Lu ng/kg 0.53 0.41 0.48 0.19 0.75 0.63

Laanthr ng/kg 2.51 1.35 1.63 1.56 ‐0.3 ‐0.6

Lanat ng/kg 2.54 1.52 1.66 1.17 2.43 2.57

Gdanthr ng/kg 2.03 24.5 58.8 13.6 10.4 11.9

Gdnat ng/kg 2.93 1.02 1.6 0.68 3.37 3.87 ΣREE ng/kg 30.3 40 78.4 25.4 40 42.8 ΣREE (Natural) ng/kg 25.7 14.1 18 10.3 30 31.4 (La + Gd) / ΣREE % 15 65 77 59 25 26 * LaSN / La SN ‐ 1.99 1.89 1.98 2.33 0.86 0.76 * CeSN / Ce SN ‐ 0.34 0.53 0.65 0.62 0.36 0.35 * EuSN / Eu SN ‐ 1.04 1.27 0.98 1.21 1.00 1.00 * GdSN / Gd SN ‐ 1.69 25.00 37.70 21.00 4.08 4.08

NdSN/YbSN ‐ 0.19 0.12 0.13 0.17 0.13 0.16

TbSN/YbSN ‐ 0.48 0.29 0.31 0.49 0.43 0.50

192

Appendix 6 continued.

Sample Lippe Type Tributary Rhine‐km 814.6 Date 30/05/2008 Sampling Campaign May 08 Weekday Fri Time 13:20 Latitude 51.64398 Longitude 6.63348 Temperature oC 20.3 pH 7.86 EC µS/cm ‐ discharge m3/s 25.1 Na+ mg/kg ‐ Mg2+ mg/kg ‐ Si mg/kg ‐ Cl‐ mg/kg ‐ K+ mg/kg ‐ Ca2+ mg/kg ‐ ‐ HCO3 mg/kg ‐ ‐ NO3 mg/kg ‐ Br‐ mg/kg ‐ 2‐ SO4 mg/kg ‐ Rb µg/kg 18.8 Sr mg/kg 1.56 Y ng/kg 23.5 Ba µg/kg 168 U µg/kg 0.57 La ng/kg 2.9 Ce ng/kg 2.56 Pr ng/kg 0.63 Nd ng/kg 3.24 Sm ng/kg 0.86 Eu ng/kg 0.23 Gd ng/kg 175 Tb ng/kg 0.22 Dy ng/kg 1.56 Ho ng/kg 0.48 Er ng/kg 1.99 Tm* ng/kg 0.37 Yb ng/kg 3.54 Lu ng/kg 0.68

Laanthr ng/kg 1.12

Lanat ng/kg 1.77

Gdanthr ng/kg 173

Gdnat ng/kg 1.16 ΣREE ng/kg 194 ΣREE (Natural) ng/kg 19.3 (La + Gd) / ΣREE % 90 * LaSN / La SN ‐ 1.63 * CeSN / Ce SN ‐ 0.54 * EuSN / Eu SN ‐ 1.08 * GdSN / Gd SN ‐ 150.00

NdSN/YbSN ‐ 0.08

TbSN/YbSN ‐ 0.22

193 Appendix 7 – May 2009: sample details and concentrations for the tributaries to the Rhine River

Sample Thur Töss Wutach Aare Wiese Elz Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 66.6 73.4 101 103.2 169.2 267.5 Date 03/05/2009 03/05/2009 04/05/2009 04/05/2009 04/05/2009 04/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Sun Sun Mon Mon Mon Mon Time 17:15 17:45 09:30 11:00 12:25 14:15 Latitude 47.59429 47.54998 47.61944 47.60036 47.58187 48.29868 Longitude 8.60689 8.55497 8.25878 8.22022 7.59182 7.74147 Temperature oC 14.4 15.5 11.6 14.3 11.9 16.1 pH 8.41 8.6 7.73 8.11 7.76 7.97 EC µS/cm 363 545 547 342 145 195 discharge m3/s 55.6 3.85 6.94 483 7.66 ‐ Na+ mg/kg 5.14 9.1 7.5 3.32 3.79 4.28 Mg2+ mg/kg 4.86 11.1 8.25 3.1 1.54 2.25 Si mg/kg 0.56 0.68 1.38 0.6 2.11 2.47 Cl‐ mg/kg 10.3 22.4 15.3 6.3 6.15 9.96 K+ mg/kg 1 1.69 1.46 0.83 0.69 0.92 Ca2+ mg/kg 27.1 38.2 40.4 28.2 9.11 12.2 ‐ HCO3 mg/kg ‐‐‐‐‐‐ ‐ NO3 mg/kg 6.28 17.6 13.3 4.7 4.36 7.38 Br‐ mg/kg 0.45 2.69 1.15 0.38 0.26 0.46 2‐ SO4 mg/kg 7.36 13.4 55.7 18.1 8.51 10.7 Rb µg/kg 1.75 2.09 3.83 1.41 2.92 2.56 Sr mg/kg 0.19 0.27 0.41 0.32 0.08 0.08 Y ng/kg 3.21 0.89 3.37 2.53 29.5 12.3 Ba µg/kg 30.8 51.7 80.9 31.4 58 58.6 U µg/kg 0.44 1.01 0.59 0.52 0.19 0.2 La ng/kg 2.62 0.76 6.27 1.94 26.9 10.1 Ce ng/kg 4.43 1.83 6.7 2.91 28.7 11.5 Pr ng/kg 0.89 0.27 2.13 0.55 7.49 2.73 Nd ng/kg 4.15 1.27 9.93 2.71 33.4 12.6 Sm ng/kg 1.08 0.36 2.76 0.65 7.81 3.41 Eu ng/kg 0.28 0.1 0.71 0.17 1.73 1.06 Gd ng/kg 2.31 1.21 5.63 1.24 8.9 5.03 Tb ng/kg 0.22 0.07 0.45 0.14 1.1 0.59 Dy ng/kg 1.45 0.48 2.75 0.87 6.47 3.52 Ho ng/kg 0.3 0.11 0.58 0.2 1.39 0.77 Er ng/kg 0.95 0.39 1.78 0.66 4.31 2.26 Tm* ng/kg 0.14 0.06 0.25 0.1 0.67 0.34 Yb ng/kg 0.98 0.48 1.77 0.71 5.13 2.57 Lu ng/kg 0.15 0.08 0.31 0.12 0.96 0.46

Laanthr ng/kg 0.27 0.14 1.21 0.19 4.86 3.49

Lanat ng/kg 2.35 0.62 5.06 1.74 22.1 6.65

Gdanthr ng/kg 0.88 0.68 1.69 0.45 ‐0.5 0.29

Gdnat ng/kg 1.43 0.53 3.94 0.79 9.37 4.73 ΣREE ng/kg 19.9 7.5 42 13 135 57 ΣREE (Natural) ng/kg 18.8 6.68 39.1 12.3 131 53.2 (La + Gd) / ΣREE % 5.8 11 6.9 5 3.2 6.6 * LaSN / La SN ‐ 1.12 1.23 1.24 1.11 1.22 1.52 * CeSN / Ce SN ‐ 0.72 1.07 0.49 0.66 0.52 0.65 * EuSN / Eu SN ‐ 1.07 1.04 1.01 1.09 0.95 1.25 * GdSN / Gd SN ‐ 1.61 2.26 1.43 1.58 0.95 1.06

NdSN/YbSN ‐ 0.35 0.22 0.47 0.32 0.54 0.41

TbSN/YbSN ‐ 0.81 0.56 0.92 0.69 0.78 0.84

194 Appendix 7 continued.

Sample Kinzig Ill Acher/Rench Moder Sauer Murg Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 299.4 312 315.3 334.4 344 344.6 Date 04/05/2009 04/05/2009 04/05/2009 05/05/2009 05/05/2009 05/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Mon Mon Mon Tue Tue Tue Time 16:45 17:37 17:55 09:15 09:50 11:30 Latitude 48.57573 48.68879 48.68271 48.80944 48.8979 48.9035 Longitude 7.82399 7.90925 7.97675 8.05641 8.11563 8.18033 Temperature oC 16.2 17.1 18 12.9 12.7 12.1 pH 7.71 8 8.41 7.65 7.69 7.62 EC µS/cm 220 452 398 506 479 153 discharge m3/s 11.2 ‐‐‐‐ 7.27 Na+ mg/kg 6.53 9.23 5.51 9.56 9.96 5.55 Mg2+ mg/kg 2.44 4.49 4.2 6.04 4.87 1.28 Si mg/kg 1.62 1.3 1.97 1.69 1.56 1.28 Cl‐ mg/kg 16.3 30.1 12.9 28.8 39.9 17.7 K+ mg/kg 1.53 1.46 1.37 2.17 1.81 1.13 Ca2+ mg/kg 13.3 34 33.3 36.2 35.6 8.1 ‐ HCO3 mg/kg ‐‐‐‐‐‐ ‐ NO3 mg/kg 5.99 7.01 8.15 6.48 12.6 5.61 Br‐ mg/kg 0.52 0.42 0.44 0.44 0.09 0.23 2‐ SO4 mg/kg 12.6 25.4 20.2 47.4 30.6 18.6 Rb µg/kg 6.49 2.19 4.25 3.47 4.75 4.79 Sr mg/kg 0.08 0.34 0.17 0.41 0.36 0.04 Y ng/kg 11.3 5.95 7.48 4.95 5.7 50.5 Ba µg/kg 138 72.9 73.1 68.3 51.7 105 U µg/kg 0.3 0.8 0.74 0.75 0.63 0.1 La ng/kg 8.18 7.23 4.4 7.12 6.91 17.3 Ce ng/kg 8.79 8.05 6.25 11.7 9.19 22.3 Pr ng/kg 2.45 1.99 1.04 2.12 1.95 5.34 Nd ng/kg 12.3 9.32 4.64 10.1 9.11 26.7 Sm ng/kg 3.39 2.49 1.71 2.52 2.46 8.86 Eu ng/kg 0.79 0.56 0.51 0.63 0.6 2.24 Gd ng/kg 5.78 4.26 6.3 3.93 4.11 17.2 Tb ng/kg 0.57 0.45 0.43 0.48 0.47 1.86 Dy ng/kg 3.74 3.41 3.06 3.27 3.25 11.1 Ho ng/kg 0.78 0.79 0.72 0.8 0.78 2.33 Er ng/kg 2.44 2.79 2.52 3.12 2.62 7.07 Tm* ng/kg 0.37 0.45 0.39 0.53 0.42 0.99 Yb ng/kg 2.71 3.54 2.98 4.58 3.39 6.97 Lu ng/kg 0.48 0.69 0.52 0.91 0.62 1.12

Laanthr ng/kg 1.88 2.19 2.85 1.04 2.07 6.81

Lanat ng/kg 6.3 5.05 1.55 6.08 4.84 10.5

Gdanthr ng/kg 0.97 0.85 3.07 0.72 0.71 2.16

Gdnat ng/kg 4.81 3.41 3.23 3.22 3.4 15 ΣREE ng/kg 52.7 46 35.5 51.9 45.9 131 ΣREE (Natural) ng/kg 49.9 43 29.5 50.1 43.1 122 (La + Gd) / ΣREE % 5.4 6.6 17 3.4 6 6.8 * LaSN / La SN ‐ 1.30 1.43 2.84 1.17 1.43 1.65 * CeSN / Ce SN ‐ 0.52 0.60 1.29 0.75 0.71 0.72 * EuSN / Eu SN ‐ 0.92 0.91 1.02 1.03 0.98 0.92 * GdSN / Gd SN ‐ 1.20 1.25 1.95 1.22 1.21 1.14

NdSN/YbSN ‐ 0.38 0.22 0.13 0.18 0.22 0.32

TbSN/YbSN ‐ 0.77 0.46 0.53 0.38 0.51 0.97

195

Appendix 7 continued.

Sample Lauter Neckar Pfrimm Main Nahe Lahn Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 355.6 428.2 446.7 496.7 529.2 585.8 Date 05/05/2009 05/05/2009 05/05/2009 06/05/2009 06/05/2009 06/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Tue Tue Tue Wed Wed Wed Time 10:35 16:45 17:50 12:10 14:05 15:20 Latitude 48.98192 49.49463 49.66065 50.00068 49.9669 50.30885 Longitude 8.21247 8.46724 8.36579 8.41915 7.89106 7.60545 Temperature oC 11.8 15.3 11.8 15.7 15.1 14.3 pH 7.61 7.94 8.3 7.97 7.98 8.41 EC µS/cm 211 828 912 704 517 445 discharge m3/s ‐ 93.1 ‐ 166 9.89 20.6 Na+ mg/kg 4.42 15.1 16.6 15.2 18.1 13.1 Mg2+ mg/kg 3.06 11 15.8 10.7 7.65 6.97 Si mg/kg 2.07 0.81 3.06 1.5 1.35 1.73 Cl‐ mg/kg 11.3 47.6 72.1 45.9 54.9 45.3 K+ mg/kg 1.88 2.46 5.22 2.74 3.77 2.35 Ca2+ mg/kg 12 60.2 57.2 41.1 22.9 23.9 ‐ HCO3 mg/kg ‐ ‐‐‐‐ ‐ ‐ NO3 mg/kg 4.28 15.6 31.4 15.9 10.4 8.27 Br‐ mg/kg 0.2 0.66 0.72 0.53 0.09 0.48 2‐ SO4 mg/kg 12.6 120 78.2 83 29.6 27.1 Rb µg/kg 4.35 3.97 4.98 4 5.11 4.15 Sr mg/kg 0.06 0.74 0.57 0.56 0.26 0.16 Y ng/kg 7.98 3.28 5.58 4.43 3.57 2.29 Ba µg/kg 48.5 108 151 74.7 79.8 26.8 U µg/kg 0.14 0.93 6.12 0.92 1.24 0.27 La ng/kg 9.81 5.43 8.85 5.48 8.24 6.26 Ce ng/kg 13.3 6.06 12.7 6.43 5.73 4 Pr ng/kg 2.31 1.5 2.95 2.08 1.6 0.96 Nd ng/kg 9.98 7.51 14.6 11 7.38 4.22 Sm ng/kg 2.47 2.02 4.32 3.31 1.95 1.05 Eu ng/kg 0.62 0.48 0.99 0.87 0.49 0.28 Gd ng/kg 3.86 23.2 23.6 32 32 24.6 Tb ng/kg 0.43 0.37 0.85 0.69 0.34 0.17 Dy ng/kg 2.64 2.69 5.73 4.15 2.42 1.14 Ho ng/kg 0.57 0.68 1.35 0.91 0.58 0.28 Er ng/kg 1.82 2.1 4.17 2.86 2.13 0.97 Tm* ng/kg 0.29 0.3 0.6 0.42 0.35 0.16 Yb ng/kg 2.25 2.12 4.26 3.02 2.84 1.36 Lu ng/kg 0.4 0.34 0.69 0.53 0.55 0.25

Laanthr ng/kg 3.75 1.4 2.1 0.47 4.18 3.71

Lanat ng/kg 6.06 4.03 6.76 5.01 4.06 2.54

Gdanthr ng/kg 0.73 20.4 17 26.9 29.3 23.2

Gdnat ng/kg 3.13 2.78 6.57 5.09 2.64 1.33 ΣREE ng/kg 50.8 54.8 85.6 73.8 66.6 45.7 ΣREE (Natural) ng/kg 46.3 33 66.5 46.4 33.1 18.7 (La + Gd) / ΣREE % 8.8 40 22 37 50 59 * LaSN / La SN ‐ 1.62 1.35 1.31 1.09 2.03 2.46 * CeSN / Ce SN ‐ 0.86 0.56 0.67 0.46 0.53 0.61 * EuSN / Eu SN ‐ 1.05 0.96 0.88 1.00 1.02 1.10 * GdSN / Gd SN ‐ 1.23 8.33 3.59 6.30 12.10 18.40

NdSN/YbSN ‐ 0.37 0.29 0.29 0.30 0.22 0.26

TbSN/YbSN ‐ 0.70 0.64 0.73 0.83 0.43 0.46

196

Appendix 7 continued.

Sample Mosel Wied Ahr Sieg Wupper Wupper Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 592.3 610.3 629.3 659.5 703.4 703.4 Date 06/05/2009 06/05/2009 06/05/2009 06/05/2009 07/05/2009 07/05/2009 Sampling Campaign May 09 May 09 May 09 May 09 May 09 May 09 Weekday Wed Wed Wed Wed Thu Thu Time 16:35 17:10 18:25 19:15 08:50 09:02 Latitude 50.3631 50.44075 50.54976 50.76556 51.04565 51.04623 Longitude 7.60081 7.45165 7.24459 7.10716 6.94844 6.94941 Temperature oC 15.3 12 12.3 12.2 12.2 12.2 pH 8.58 7.8 7.55 7.69 7.73 7.71 EC µS/cm 952 278 505 333 402 429 discharge m3/s 123 6.22 2.8 29.3 7.56 7.56 Na+ mg/kg 28.2 9.13 9.63 10.3 13.3 15.6 Mg2+ mg/kg 10 3.91 8.66 3.77 4.29 4.31 Si mg/kg 0.44 1.51 0.83 1.04 1.16 1.14 Cl‐ mg/kg 129 31.4 47.2 25.8 26 35 K+ mg/kg 2.47 1.71 1.66 2.13 2.43 2.75 Ca2+ mg/kg 51.6 11.2 25.1 15.8 18.9 18.6 ‐ HCO3 mg/kg ‐‐‐‐‐ ‐ ‐ NO3 mg/kg 7.81 10 16.9 10.4 11.7 12.8 Br‐ mg/kg 0.32 0.21 0.73 0.25 0.18 ‐ 2‐ SO4 mg/kg 65.6 20.2 35.9 20.6 24.5 34 Rb µg/kg 4.93 2.6 2.03 4.08 3.64 3.96 Sr mg/kg 0.67 0.09 0.15 0.11 0.09 0.09 Y ng/kg 1.95 4.37 2.22 3.57 2.35 1.83 Ba µg/kg 55.4 20.1 19.9 21.4 22.2 20.2 U µg/kg 0.56 0.06 0.23 0.06 0.09 0.06 La ng/kg 3.96 9.29 3.43 5.24 3.93 2.8 Ce ng/kg 3.51 9.61 2.15 5.25 3 2.24 Pr ng/kg 1.1 1.94 0.79 1.45 0.89 0.71 Nd ng/kg 5.44 7.99 3.95 6.83 4.32 3.45 Sm ng/kg 1.47 1.95 1.19 1.84 1.20 0.94 Eu ng/kg 0.38 0.49 0.31 0.47 0.29 0.23 Gd ng/kg 6.79 13.9 5.21 29.8 72.7 84.1 Tb ng/kg 0.29 0.33 0.24 0.32 0.23 0.19 Dy ng/kg 1.95 1.98 1.47 1.98 1.76 1.42 Ho ng/kg 0.48 0.46 0.37 0.43 0.43 0.37 Er ng/kg 1.72 1.36 1.34 1.55 1.49 1.3 Tm* ng/kg 0.27 0.21 0.21 0.26 0.33 0.29 Yb ng/kg 2.19 1.55 1.7 2.2 3.39 2.96 Lu ng/kg 0.4 0.28 0.34 0.43 0.83 0.61

Laanthr ng/kg 1.08 4.35 1.64 1.58 1.74 0.98

Lanat ng/kg 2.88 4.95 1.8 3.66 2.2 1.82

Gdanthr ng/kg 4.74 11.4 3.39 27.3 71 82.8

Gdnat ng/kg 2.05 2.44 1.82 2.53 1.71 1.3 ΣREE ng/kg 29.9 51.3 22.7 58.1 94.8 102 ΣREE (Natural) ng/kg 24.1 35.5 17.7 29.2 22.1 17.8 (La + Gd) / ΣREE % 19 31 22 50 77 82 * LaSN / La SN ‐ 1.37 1.88 1.91 1.43 1.79 1.54 * CeSN / Ce SN ‐ 0.45 0.76 0.42 0.54 0.50 0.46 * EuSN / Eu SN ‐ 1.04 1.05 1.00 1.02 0.96 0.99 * GdSN / Gd SN ‐ 3.31 5.69 2.86 11.80 42.40 64.40

NdSN/YbSN ‐ 0.21 0.43 0.19 0.26 0.11 0.10

TbSN/YbSN ‐ 0.49 0.77 0.51 0.53 0.25 0.23

197

Appendix 7 continued.

Sample Erft Ruhr Lippe Type Tributary Tributary Tributary Rhine‐km 735.6 780.7 814.6 Date 07/05/2009 07/05/2009 07/05/2009 Sampling Campaign May 09 May 09 May 09 Weekday Thu Thu Thu Time 10:10 12:15 13:37 Latitude 51.18269 51.44582 51.64398 Longitude 6.73103 6.72884 6.63348 Temperature oC 17.5 16.2 15.3 pH 7.85 7.69 7.97 EC µS/cm 929 565 1837 discharge m3/s 8.5 35.9 23.2 Na+ mg/kg 44.4 23 112 Mg2+ mg/kg 11.7 5.11 7.05 Si mg/kg 3.9 0.25 2.04 Cl‐ mg/kg 124 57.2 522 K+ mg/kg 6.15 2.82 4.53 Ca2+ mg/kg 31.9 24.3 67.9 ‐ HCO3 mg/kg ‐ ‐ ‐ ‐ NO3 mg/kg 12 10.3 42.5 Br‐ mg/kg 0.11 0.32 0.64 2‐ SO4 mg/kg 107 47.7 120 Rb µg/kg 22.9 4.63 10.8 Sr mg/kg 0.40 0.22 0.02 Y ng/kg 1.15 2.53 0.92 Ba µg/kg 133 38.4 260 U µg/kg 0.36 0.14 0.6 La ng/kg 3.09 2.36 4.58 Ce ng/kg 2.61 2.25 4.31 Pr ng/kg 0.56 0.91 0.96 Nd ng/kg 2.53 4.99 4.66 Sm ng/kg 0.67 1.49 1.14 Eu ng/kg 0.18 0.42 0.31 Gd ng/kg 20.2 29.5 314 Tb ng/kg 0.15 0.33 0.27 Dy ng/kg 1.09 2.35 1.91 Ho ng/kg 0.27 0.58 0.53 Er ng/kg 1 1.88 1.84 Tm* ng/kg 0.16 0.38 0.37 Yb ng/kg 1.31 3.54 3.61 Lu ng/kg 0.32 0.62 0.68

Laanthr ng/kg 1.72 0.07 1.71

Lanat ng/kg 1.37 2.29 2.87

Gdanthr ng/kg 19.3 27.2 313

Gdnat ng/kg 0.92 2.27 1.43 ΣREE ng/kg 34.1 51.6 339 ΣREE (Natural) ng/kg 13.1 24.3 24.9 (La + Gd) / ΣREE % 62 53 93 * LaSN / La SN ‐ 2.25 1.03 1.60 * CeSN / Ce SN ‐ 0.71 0.35 0.59 * EuSN / Eu SN ‐ 1.10 1.09 1.13 * GdSN / Gd SN ‐ 22.10 13.00 219.00

NdSN/YbSN ‐ 0.16 0.12 0.11

TbSN/YbSN ‐ 0.43 0.34 0.28

198 Appendix 8 – October 2009: sample details and concentrations for the tributaries to the Rhine River

Sample Thur Töss Wutach Aare Wiese Elz Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 66.6 73.4 101 103.2 169.2 267.5 Date 25/10/2009 25/10/2009 25/10/2009 25/10/2009 25/10/2009 25/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Sun Sun Sun Sun Sun Sun Time 09:40 10:43 11:53 12:56 15:10 17:25 Latitude 47.59429 47.54998 47.61944 47.60036 47.58187 48.29868 Longitude 8.60689 8.55497 8.25878 8.22022 7.59182 7.74147 Temperature oC 10 11.8 10.8 14.1 12 12.1 pH 8.12 8.22 7.83 7.98 8.42 7.8 EC µS/cm 445 690 626 318 215 260 discharge m3/s 27.4 1.93 2.65 309 2.68 ‐ Na+ mg/kg 7.2 19.2 12.2 4.56 8.68 7.21 Mg2+ mg/kg 10.3 21.1 16.2 5.51 3.86 5.34 Si mg/kg 0.78 1.02 1.62 0.57 1.97 2.82 Cl‐ mg/kg 26.3 83.2 50.7 15 26.2 27.1 K+ mg/kg 3.35 7 4.22 2.27 3.17 3.22 Ca2+ mg/kg 60.4 74.3 72.3 41.8 21.3 27.2 ‐ HCO3 mg/kg 164 123 42.3 78.4 31.1 49.7 ‐ NO3 mg/kg 13 53.4 29.7 8.43 9.13 13.2 Br‐ mg/kg 1.77 1.51 1.64 1.16 0.72 0.9 2‐ SO4 mg/kg 15.9 39.9 174 34.2 28.9 30.1 Rb µg/kg 1.4 3.01 4.1 1.48 4.49 2.74 Sr mg/kg 0.24 0.27 0.48 0.3 0.1 0.11 Y ng/kg 16.9 5.26 17 7.34 34.3 31.5 Ba µg/kg 41.9 54.8 85.1 31.1 79.4 71.9 U µg/kg 0.46 1.19 0.77 0.58 0.52 0.42 La ng/kg 3.71 0.98 5.92 1.6 21.5 15.6 Ce ng/kg 4.72 1.67 6.07 2.05 24.4 18.3 Pr ng/kg 1.01 0.3 1.76 0.46 5.68 4.01 Nd ng/kg 4.98 1.69 8.06 2.3 25.1 18 Sm ng/kg 1.27 0.46 2.06 0.56 5.57 4.43 Eu ng/kg 0.3 0.11 0.55 0.15 1.21 1.34 Gd ng/kg 5.08 42.8 11.9 7.06 9.97 11.3 Tb ng/kg 0.31 0.11 0.33 0.14 0.79 0.78 Dy ng/kg 1.86 0.7 2.17 0.85 5.1 4.68 Ho ng/kg 0.42 0.17 0.5 0.17 1.21 1.06 Er ng/kg 1.4 0.51 1.56 0.55 4.45 3.19 Tm* ng/kg 0.2 0.09 0.24 0.09 0.76 0.5 Yb ng/kg 1.51 0.75 1.75 0.68 6.56 3.82 Lu ng/kg 0.38 0.18 0.37 0.14 1.49 0.81

Laanthr ng/kg 0.85 0.1 1.27 0.16 3.6 4.49

Lanat ng/kg 2.86 0.88 4.66 1.44 17.9 11.1

Gdanthr ng/kg 3.4 42.2 9.16 6.37 3.61 5.7

Gdnat ng/kg 1.68 0.65 2.7 0.69 6.36 5.57 ΣREE ng/kg 27.1 50.6 43.2 16.8 114 87.8 ΣREE (Natural) ng/kg 22.9 8.27 32.8 10.3 106 77.6 (La + Gd) / ΣREE % 16 84 24 39 6.3 12 * LaSN / La SN ‐ 1.30 1.11 1.27 1.11 1.20 1.40 * CeSN / Ce SN ‐ 0.63 0.71 0.50 0.56 0.56 0.65 * EuSN / Eu SN ‐ 0.97 0.98 1.09 1.16 0.96 1.27 * GdSN / Gd SN ‐ 3.03 66.00 4.39 10.20 1.57 2.02

NdSN/YbSN ‐ 0.27 0.19 0.38 0.28 0.32 0.39

TbSN/YbSN ‐ 0.75 0.52 0.69 0.76 0.44 0.74

199

Appendix 8 continued.

Sample Kinzig Ill Acher/Rench Moder Sauer Murg Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 299.4 312 315.3 334.4 344 344.6 Date 26/10/2009 26/10/2009 26/10/2009 26/10/2009 26/10/2009 26/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Mon Mon Mon Mon Mon Mon Time 10:18 11:11 11:41 12:20 13:01 14:35 Latitude 48.57573 48.68879 48.68271 48.80944 48.8979 48.9035 Longitude 7.82399 7.90925 7.97675 8.05641 8.11563 8.18033 Temperature oC 12 12.4 12.6 11.8 10.9 12.2 pH 7.62 7.94 7.16 7.54 7.76 8.13 EC µS/cm 273 504 247 486 533 193 discharge m3/s 6.08 ‐ ‐ ‐ ‐ 6.14 Na+ mg/kg 15 15.1 12.3 32.7 23 9.95 Mg2+ mg/kg 4.3 8.52 3.84 9.54 8.64 2.58 Si mg/kg 1.83 1.36 1.93 2.04 2.41 1.42 Cl‐ mg/kg 54.3 88.4 34.9 83.9 74.4 17.5 K+ mg/kg 6.34 4.59 6.36 6.86 8.71 3.85 Ca2+ mg/kg 28.8 58.8 21.1 49.1 46.8 15.5 ‐ HCO3 mg/kg 35.7 60.2 31.6 53.1 79.9 31.2 ‐ NO3 mg/kg 13.3 14.1 14.2 10.2 9.49 4.31 Br‐ mg/kg 0.73 0.81 0.34 0.75 0.73 0.09 2‐ SO4 mg/kg 35.1 60.5 33.6 93.9 52.1 15.8 Rb µg/kg 7.69 1.91 7.15 3.79 13.3 5.48 Sr mg/kg 0.08 0.39 0.07 0.36 0.38 0.05 Y ng/kg 34.7 27.3 48 39.1 22.7 94.1 Ba µg/kg 141 73 78.4 67.4 52.9 91.6 U µg/kg 0.26 0.91 0.17 0.62 0.36 0.11 La ng/kg 12.9 5.46 5.17 5.45 6.45 16.2 Ce ng/kg 16.1 5.55 7.5 11.9 10.3 24.6 Pr ng/kg 3.7 1.5 1.68 2.09 1.79 6.64 Nd ng/kg 17.4 7.35 7.94 9.88 8.16 33.3 Sm ng/kg 4.77 2.05 2.98 3.06 2.04 11.1 Eu ng/kg 1.14 0.51 0.9 0.8 0.52 2.82 Gd ng/kg 28.5 9.77 18.4 16.6 8 39.7 Tb ng/kg 0.83 0.43 0.91 0.73 0.45 2.21 Dy ng/kg 5.24 3.22 6.88 5.36 2.99 13.8 Ho ng/kg 1.15 0.78 1.76 1.33 0.72 2.95 Er ng/kg 3.45 2.76 6.21 4.76 2.37 8.61 Tm* ng/kg 0.53 0.45 0.97 0.78 0.38 1.23 Yb ng/kg 3.93 3.7 7.55 6.43 3.05 8.61 Lu ng/kg 0.62 0.78 1.35 1.33 0.63 1.39

Laanthr ng/kg 3.87 1.72 2.58 1.16 1.59 3.34

Lanat ng/kg 9.07 3.74 2.58 4.29 4.85 12.9

Gdanthr ng/kg 21.9 6.85 12.7 11.8 5.37 20.7

Gdnat ng/kg 6.69 2.92 5.72 4.85 2.63 19 ΣREE ng/kg 100 44.3 70.2 70.5 47.8 173 ΣREE (Natural) ng/kg 74.6 35.8 54.9 57.6 40.8 149 (La + Gd) / ΣREE % 26 19 22 18 15 14 * LaSN / La SN ‐ 1.43 1.46 2.00 1.27 1.33 1.26 * CeSN / Ce SN ‐ 0.66 0.55 0.92 0.97 0.82 0.64 * EuSN / Eu SN ‐ 0.95 0.99 1.03 0.97 1.05 0.91 * GdSN / Gd SN ‐ 4.27 3.35 3.22 3.43 3.04 2.09

NdSN/YbSN ‐ 0.37 0.17 0.09 0.13 0.22 0.32

TbSN/YbSN ‐ 0.77 0.43 0.44 0.42 0.54 0.94

200

Appendix 8 continued.

Sample Lauter Neckar Pfrimm Main Nahe Lahn Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 355.6 428.2 446.7 496.7 529.2 585.8 Date 26/10/2009 26/10/2009 27/10/2009 27/10/2009 27/10/2009 27/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Mon Mon Tue Tue Tue Tue Time 13:48 17:05 09:20 13:04 15:07 16:38 Latitude 48.98192 49.49463 49.66065 50.00068 49.9669 50.30885 Longitude 8.21247 8.46724 8.36579 8.41915 7.89106 7.60545 Temperature oC 10.6 11.1 11 12.3 11.3 9.6 pH 7.6 7.85 8.2 7.78 8.26 7.6 EC µS/cm 193 830 1180 870 635 507 discharge m3/s ‐ 54.1 ‐ 107 6.26 12.6 Na+ mg/kg 6.02 23.1 37.5 32.1 31.8 20.4 Mg2+ mg/kg 4.85 19.2 33.3 22.3 14.3 12.6 Si mg/kg 2.34 1.7 2.97 2.21 1.67 3.19 Cl‐ mg/kg 29.3 60.9 133 82.5 89.9 46.9 K+ mg/kg 5.62 8.63 25.2 10.4 14.4 7.52 Ca2+ mg/kg 20.6 94.5 108 86.9 41.3 40.5 ‐ HCO3 mg/kg 31.4 128 197 128 76.3 112 ‐ NO3 mg/kg 7.76 18.8 47.6 17.3 13.4 11.9 Br‐ mg/kg 0.62 0.65 0.1 0.12 0.11 0.05 2‐ SO4 mg/kg 23.5 141 116 115 39.2 26.6 Rb µg/kg 4.53 4.78 9.89 6.09 7.73 5.95 Sr mg/kg 0.05 0.79 0.69 0.65 0.30 0.27 Y ng/kg 11.2 23.3 40.2 25.5 16.1 10.3 Ba µg/kg 51.4 87.1 136 73.4 82.1 35.6 U µg/kg 0.08 0.8 5.95 1.12 1.56 0.36 La ng/kg 4.58 4.48 7.86 6.01 5.21 7.66 Ce ng/kg 6.96 4.78 7.96 5.14 5.62 8.11 Pr ng/kg 1.19 1.45 2.4 1.53 1.38 1.22 Nd ng/kg 5.03 7.44 12.7 8.19 6.16 4.93 Sm ng/kg 1.38 1.97 3.78 2.41 1.60 1.16 Eu ng/kg 0.33 0.5 0.93 0.67 0.39 0.31 Gd ng/kg 19.9 53.8 25.5 73.7 55.4 42.4 Tb ng/kg 0.25 0.43 0.84 0.51 0.34 0.21 Dy ng/kg 1.49 2.68 5.12 3.33 2.11 1.25 Ho ng/kg 0.39 0.63 1.18 0.8 0.56 0.31 Er ng/kg 1.25 2.04 3.76 2.6 2.2 1.19 Tm* ng/kg 0.21 0.3 0.54 0.41 0.39 0.22 Yb ng/kg 1.75 2.21 3.91 3.21 3.62 2.05 Lu ng/kg 0.43 0.47 0.71 0.55 0.72 0.43

Laanthr ng/kg 1.96 0.4 2.03 2.18 1.74 4.43

Lanat ng/kg 2.62 4.08 5.82 3.83 3.47 3.23

Gdanthr ng/kg 17.9 51.2 19.7 70 53.3 41

Gdnat ng/kg 1.94 2.67 5.77 3.63 2.13 1.4 ΣREE ng/kg 45.1 83.2 77.1 109 85.7 71.5 ΣREE (Natural) ng/kg 25.2 31.7 55.4 36.8 30.7 26 (La + Gd) / ΣREE % 44 62 28 66 64 64 * LaSN / La SN ‐ 1.75 1.10 1.35 1.57 1.50 2.37 * CeSN / Ce SN ‐ 0.99 0.44 0.49 0.48 0.62 1.01 * EuSN / Eu SN ‐ 0.94 1.03 0.94 1.07 0.99 1.15 * GdSN / Gd SN ‐ 10.30 20.20 4.42 20.30 26.00 30.40

NdSN/YbSN ‐ 0.24 0.28 0.27 0.21 0.14 0.20

TbSN/YbSN ‐ 0.51 0.71 0.78 0.58 0.34 0.37

201 Appendix 8 continued.

Sample Mosel Wied Ahr Sieg Wupper Erft Type Tributary Tributary Tributary Tributary Tributary Tributary Rhine‐km 592.3 610.3 629.3 659.5 703.4 735.6 Date 27/10/2009 28/10/2009 28/10/2009 28/10/2009 28/10/2009 28/10/2009 Sampling Campaign Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Oct 09 Weekday Tue Wed Wed Wed Wed Wed Time 17:29 09:02 10:19 12:33 14:27 15:31 Latitude 50.3631 50.44075 50.5528 50.76556 51.04565 51.18269 Longitude 7.60081 7.45165 7.26491 7.10716 6.94844 6.73103 Temperature oC 12 9.3 11.5 11 12.6 19 pH 7.78 7.68 7.56 7.9 8.03 7.87 EC µS/cm 1218 361 718 382 440 876 discharge m3/s 67.2 1.95 0.798 12.9 6.51 7.8 Na+ mg/kg 58.8 18.1 17.9 18.1 21.7 62.3 Mg2+ mg/kg 22.6 8.35 19.2 7.77 8.19 19 Si mg/kg 1.54 1.82 0.65 1.4 1.26 4.38 Cl‐ mg/kg 247 42.4 42.4 41 46.5 66.1 K+ mg/kg 10.8 6.81 6.42 7.6 7.55 16 Ca2+ mg/kg 102 24.7 49.6 28.7 32.6 44.4 ‐ HCO3 mg/kg 61.7 49 162 55.9 55.8 193 ‐ NO3 mg/kg 11.3 7.44 8.05 11.7 12.1 2.15 Br‐ mg/kg 0.4 0.27 0.66 0.28 0.27 0.44 2‐ SO4 mg/kg 167 24.6 28.7 24.9 42.7 57.8 Rb µg/kg 8.14 3.6 3.75 4.22 4.39 24.8 Sr mg/kg 0.97 0.10 0.16 0.10 0.09 0.44 Y ng/kg 14.7 12.1 10.4 8.41 8.42 6.63 Ba µg/kg 64.5 22.4 19.3 18.8 21.9 158 U µg/kg 0.65 0.1 0.33 0.07 0.11 0.18 La ng/kg 4.24 5.52 0.79 3.39 2.6 1.5 Ce ng/kg 2.18 5.67 1.28 4.4 2.88 2.34 Pr ng/kg 0.74 1.24 0.25 0.82 0.64 0.34 Nd ng/kg 3.67 5.41 1.26 3.68 3.11 1.58 Sm ng/kg 1.04 1.36 0.49 1.01 0.86 0.42 Eu ng/kg 0.27 0.35 0.15 0.28 0.23 0.13 Gd ng/kg 22.7 45.6 6.59 54.9 77.4 17.4 Tb ng/kg 0.21 0.25 0.14 0.2 0.17 0.11 Dy ng/kg 1.6 1.64 1.16 1.22 1.09 0.69 Ho ng/kg 0.46 0.39 0.36 0.35 0.28 0.17 Er ng/kg 1.73 1.35 1.65 1.4 1.11 0.56 Tm* ng/kg 0.31 0.22 0.3 0.27 0.22 0.09 Yb ng/kg 2.7 1.86 2.79 2.67 2.14 0.71 Lu ng/kg 0.53 0.35 0.59 0.74 0.47 0.13

Laanthr ng/kg 2.42 2.33 0.4 1.48 1.01 0.66

Lanat ng/kg 1.81 3.19 0.39 1.91 1.59 0.84

Gdanthr ng/kg 21.1 43.8 5.62 53.5 76.1 16.8

Gdnat ng/kg 1.51 1.76 0.97 1.42 1.23 0.59 ΣREE ng/kg 42.3 71.2 17.8 75.3 93.2 26.1 ΣREE (Natural) ng/kg 18.8 25.1 11.8 20.4 16 8.7 (La + Gd) / ΣREE % 56 65 34 73 83 67 * LaSN / La SN ‐ 2.34 1.73 2.03 1.78 1.63 1.78 * CeSN / Ce SN ‐ 0.44 0.69 1.03 0.85 0.67 1.04 * EuSN / Eu SN ‐ 1.01 1.07 1.00 1.10 1.05 1.25 * GdSN / Gd SN ‐ 15.00 25.90 6.79 38.60 63.10 29.60

NdSN/YbSN ‐ 0.11 0.24 0.04 0.11 0.12 0.19

TbSN/YbSN ‐ 0.29 0.49 0.18 0.28 0.29 0.57

202

Appendix 8 continued.

Sample Ruhr Lippe Type Tributary Tributary Rhine‐km 780.7 814.6 Date 28/10/2009 28/10/2009 Sampling Campaign Oct 09 Oct 09 Weekday Wed Wed Time 17:48 19:00 Latitude 51.44582 51.64398 Longitude 6.72884 6.63348 Temperature oC 11.5 7.27 pH 7.6 7.87 EC µS/cm 260 22201 discharge m3/s 29 17.5 Na+ mg/kg 19.4 236 Mg2+ mg/kg 7.4 13.2 Si mg/kg 1.48 2.26 Cl‐ mg/kg 37.7 426 K+ mg/kg 6.9 12.1 Ca2+ mg/kg 36 109 ‐ HCO3 mg/kg 41.6 79.3 ‐ NO3 mg/kg 6.06 13.3 Br‐ mg/kg 0.23 0.6 2‐ SO4 mg/kg 32.3 85.1 Rb µg/kg 4.34 1.3 Sr mg/kg 0.20 Y ng/kg 8.8 15.3 Ba µg/kg 30.9 35.9 U µg/kg 0.13 0.06 La ng/kg 8.9 10.1 Ce ng/kg 2.26 3.98 Pr ng/kg 0.58 0.78 Nd ng/kg 3.14 3.88 Sm ng/kg 1.04 0.96 Eu ng/kg 0.31 0.25 Gd ng/kg 39.6 51.7 Tb ng/kg 0.25 0.24 Dy ng/kg 1.57 1.54 Ho ng/kg 0.42 0.4 Er ng/kg 1.73 1.51 Tm* ng/kg 0.33 0.29 Yb ng/kg 3.18 2.78 Lu ng/kg 0.78 0.67

Laanthr ng/kg 7.67 7.7

Lanat ng/kg 1.23 2.36

Gdanthr ng/kg 37.8 50.5

Gdnat ng/kg 1.77 1.21 ΣREE ng/kg 64 79.1 ΣREE (Natural) ng/kg 18.6 20.9 (La + Gd) / ΣREE % 71 74 * LaSN / La SN ‐ 7.26 4.26 * CeSN / Ce SN ‐ 0.62 0.66 * EuSN / Eu SN ‐ 1.08 1.11 * GdSN / Gd SN ‐ 22.40 42.60

NdSN/YbSN ‐ 0.08 0.12

TbSN/YbSN ‐ 0.28 0.32

203 Appendix 9 – Sample details and concentrations for the Wiembach Creek and Lake Constance

Sample Lake Constance Lake Constance Wiembach Wiembach Wiembach Type Rhine Rhine Control Control Control Rhine‐km ‐38 ‐38 ‐‐‐ Date 20/08/2011 20/08/2011 29/05/2008 06/05/2009 26/10/2009 Sampling Campaign Aug 11 Aug 11 May 08 May 09 Oct 09 Weekday Sat Sat Thu Wed Mon Time ‐ ‐ 17:10 20:45 12:00 Latitude 47.48905 47.48905 51.07335 51.07335 51.07335 Longitude 9.60265 9.60265 7.00322 7.00322 7.00322 Temperature oC ‐‐ 16.5 11.2 11.9 pH 7.95 7.98 8.14 7.99 8.02 EC µS/cm 283 313 631 563 601 discharge m3/s ‐‐ ‐ ‐ ‐ Na+ mg/kg ‐‐‐7.19 ‐ Mg2+ mg/kg ‐‐‐7.47 ‐ Si mg/kg ‐‐‐2.71 ‐ Cl‐ mg/kg ‐‐‐23.6 16.5 K+ mg/kg ‐‐‐1.12 ‐ Ca2+ mg/kg ‐‐‐41.2 ‐ ‐ HCO3 mg/kg ‐‐‐‐‐ ‐ NO3 mg/kg ‐‐‐20.3 13.8 Br‐ mg/kg ‐‐‐0.56 0.42 2‐ SO4 mg/kg ‐‐‐36.2 25.8 Rb µg/kg 1.26 1.19 3.41 1.62 1.49 Sr mg/kg 0.27 0.26 0.23 0.19 0.19 Y ng/kg 6.37 6.58 16.6 18.5 14.3 Ba µg/kg 24.7 23.4 20.3 16.4 16.2 U µg/kg 1.08 1.04 0.74 0.61 0.71 La ng/kg 1.38 1.39 2.87 5.98 6.93 Ce ng/kg 1.87 1.67 3.03 6.41 10.6 Pr ng/kg 0.33 0.31 0.71 1.43 1.82 Nd ng/kg 1.57 1.43 3.42 6.66 7.47 Sm ng/kg 0.57 0.51 0.86 1.61 1.49 Eu ng/kg 0.17 0.15 0.21 0.4 0.32 Gd ng/kg 2.06 1.95 1.14 2.39 1.49 Tb ng/kg 0.16 0.15 0.17 0.3 0.23 Dy ng/kg 0.93 0.93 1.17 1.94 1.45 Ho ng/kg 0.22 0.22 0.3 0.44 0.33 Er ng/kg 0.7 0.68 1.1 1.47 1.01 Tm* ng/kg 0.11 0.11 0.16 0.22 0.16 Yb ng/kg 0.86 0.8 1.26 1.72 1.3 Lu ng/kg 0.2 0.18 0.25 0.33 0.22

Laanthr ng/kg 0.85 0.89 0.85 1.81 0.67

Lanat ng/kg 0.53 0.5 2.02 4.17 6.26

Gdanthr ng/kg 0.98 1.02 0.03 0.39 ‐

Gdnat ng/kg 1.08 0.93 1.11 2 1.52 ΣREE ng/kg 11.1 10.5 16.6 31.3 34.9 ΣREE (Natural) ng/kg 9.31 8.56 15.8 29.1 34.2 (La + Gd) / ΣREE % 16 18 5.3 7 1.8 * LaSN / La SN ‐ 2.59 2.79 1.42 1.43 1.11 * CeSN / Ce SN ‐ 1.13 1.09 0.58 0.61 0.74 * EuSN / Eu SN ‐ 1.03 1.06 1.00 1.05 1.01 * GdSN / Gd SN ‐ 1.91 2.09 1.02 1.19 0.98

NdSN/YbSN ‐ 0.15 0.15 0.23 0.32 0.48

TbSN/YbSN ‐ 0.67 0.66 0.49 0.63 0.63

204 205 Appendix 10 – General information and references for the tributaries to the Rhine River

)

3

/s)

10

3

area

(x )

(m km

(km) 2 ‐ Name

Section

R/L (km Tributary Rhine Reference Length River Station Catchment discharge population Lake 1 Constance ‐ 11,500 370 Thur 2 (Rhein) 66.6 L 135 1,696 47 High Rhine Andelfingen (2044) 3 Töss 73.4 L 56 399 9 High Rhine Freienstein

4 Wutach 101.0 R 90 130 1,252 10 High Rhine ‐ Untersiggenthal, Stilli 2 Aare 103.2 L 288 3,400 17,606 559 High Rhine (2205) Münchenstein, Hofmatt 2 Birs 164.0 L 73 911 15 High Rhine (2106) 5 Wiese 169.2 R 55 458 11 Upper Rhine Basel

6 Elz 267.5 R 90 460 1,107 22 Upper Rhine ‐

7 Kinzig 299.4 R 93 280 954 23 Upper Rhine ‐

8 Ill (Elsass) 312.0 L 217 4,600 59 Upper Rhine Chasseur‐Froid

9 Moder 334.4 L 82 1,720 17 Upper Rhine ‐

8 Sauer 344.0 L 70 541 4 Upper Rhine Beinheim

10 Murg 344.6 R 79 466 17 Upper Rhine Rotenfels

11 Lauter 355.6 L 55 382,295 3 Upper Rhine

12 Neckar 428.2 R 384 5,300 13,950 145 Upper Rhine Mannheim‐Neckar

13 Pfrimm 446.7 L 43 100 246,383 Upper Rhine

14 Main 496.7 R 553 6,600 27,251 193 Upper Rhine Frankfurt‐Osthafen

15 Nahe 529.2 L 125 4,089 32 Upper Rhine Grolsheim

15 Lahn 585.8 R 244 5,981 32 Middle Rhine Leun

16 Mosel 592.3 L 544 4,210 28,133 328 Middle Rhine ‐

15 Wied 610.3 R 102 771 8 Middle Rhine Friedrichsthal

15 Ahr 629.3 L 83 898 3 Middle Rhine Müsch

17 Sieg 659.5 R 155 2,825 53 Lower Rhine Menden

18 Wupper 703.4 R 117 827 15 Lower Rhine Opladen

19 Erft 735.6 L 103 1,828 18 Lower Rhine Neubrück

20 Ruhr 780.7 R 219 4,420 76 Lower Rhine Mülheim

21 Lippe 814.6 R 220 1,847 4,882 45 Lower Rhine Schermbeck

206 Appendix 10 continued.

Reference

1 http://www.igkb.de/html/seedaten/index.html

2 http://www.hydrodaten.admin.ch/de/

3 http://www.hw.zh.ch/hochwasser/jahrbuch/0570a007.PDF

4 http://www.rp.baden‐wuerttemberg.de/servlet/PB/show/1158622/rpf‐ref51‐wrrl‐tbg‐wutach‐18052005.pdf

5 http://www.bafu.admin.ch/publikationen/publikation/01062/index.html?lang=de

6 http://www.rp.baden‐wuerttemberg.de/servlet/PB/show/1192731/rpk52_tbg31_neu.pdf

7 http://www.rp‐karlsruhe.de/servlet/PB/show/1192733/rpk52_tbg32.pdf

8 http://www.hydro.eaufrance.fr/selection.php?consulte=rechercher

9 http://www.lorraine.developpement‐durable.gouv.fr/IMG/pdf/moder_cle212118.pdf

10 http://www.hvz.baden‐wuerttemberg.de/cgi/daten.pl?id=0111

11 http://www.lorraine.developpement‐durable.gouv.fr/IMG/pdf/lauter_cle1a7556.pdf

12 http://www.hvz.lubw.baden‐wuerttemberg.de/

13 www.luwg.rlp.de

14 http://www.hnd.bayern.de/pegel/abfluss/pegel_abfluss.php?pgnr=24088001&standalone= http://www.iksr.org/fileadmin/user_upload/Dokumente_de/Bestandsaufnahme_Teilberichte/BAG_Mittelrhe 15 in/bericht_mittelrhein_endbericht010205.pdf

16 http://www.unece.org/fileadmin/DAM/env/water/blanks/assessment/north_eastern_atlantic.pdf

17 http://luadb.lds.nrw.de/LUA/wiski/download_jb.php?name=Menden_1&jahr=2006

18 http://luadb.lds.nrw.de/LUA/wiski/download_jb.php?name=Opladen&jahr=2006

19 http://www.niederrhein.nrw.de/erft/kap_1/index.html

20 http://www.talsperrenleitzentrale‐ruhr.de/daten/internet/dokumente/dgj/q/dgj_2769990000100_q.pdf

21 http://www.niederrhein.nrw.de/lippe/index.html

207 Appendix 11 – Calculation of Rare Earth Element Anomalies

QUANTIFICATION OF ANOMALIES

As the present discussion of REE geochemistry involves anthropogenic as well as natural anomalies of several REE, the issue of quantifying the size of each anomaly will be considered here in more detail. The following section describes the motivation and justification behind the use of different methods for calculating REE anomalies in this dataset.

When calculating anomalies of REE, two principles are particularly important:

1. The natural (pristine) REE pattern that is observed for the type of sample under consideration

This principle is essential since it is the basis of the assumption made in calculating anomalies: On a semi‐logarithmic plot, normalized concentrations of elements La to Gd change smoothly, while concentrations of elements Tb to Lu also change smoothly, with a step jump from Gd to Tb separating the REE pattern in two coherent regions (Fig. 1). Thus, anomalies of elements within each of the two REE subgroups (La‐Gd and Tb‐Lu) can only be calculated using elements that belong to the same group.

2. Exclusion of (potentially) anomalous REE

In general, no REE that is anomalous or potentially anomalous should be used for calculating the size of REE anomalies. In the case of the Rhine River, the presence of several anomalies makes meeting this criterion especially difficult, but vital, as will be shown below.

Before moving onto anomalous REE, the principle of extrapolation/interpolation will be demonstrated using the element thulium (Tm). Background Tm (Tm*) concentrations have been reported here even though measured Tm concentrations are dominated by the artificially added Tm spike. Background Tm concentrations are estimated for the sake of calculating total

208

Appendix 11 continued.

dissolved REE (ΣREE), and because Tm* provides a good estimate of the background (natural) Tm concentrations. As mentioned above, the elements Tb to Lu are expected to lie on a straight line.

∗ Starting with the general equation for linear regression and substituting y for we have:

∗

* where is any given REE, SN signifies for normalization to a given standard, and signifies for interpolated/extrapolated (background) REE.

∑ 1 2 and 0 1 ∑

In calculating 1, the value for each element must be picked so that the differences in x values reflect the relative distances of the points along the x‐axis on a REE plot (e.g. = 1, = 6,

= 8 and = 14). Overbars for and signify average values, and is replaced by each REE * used (e.g. stands for the value of the REE ). In the case of Tm , would be 13.

* For Tm , Lu is excluded due to potential anomalies, and 1 is calculated using Tb, Dy, Ho, Er and Yb:

1

209 Appendix 11 continued.

2 2 2 2 2 where ,

and 5 5

The same strategy can also be used for estimating missing data, provided the REE in question and the neighboring REE used for extrapolation are not anomalous.

Toshibetsu River Delaware River Spring Creek

1 6 REE / PAAS x 10

0.1

La Ce Pr Nd Pm Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

Fig. 1. REE plots of pristine rivers showing typical increase from La towards Gd, a step‐jump from Gd to Tb, and further increase from Tb to Lu.

Natural La, Gd and Lu anomalies in natural waters can be observed due to differences between the stabilities of chemical REE complexes (e.g., Bau, 1999; Schijf and Byrne, 2001). In contrast to these small positive natural anomalies, the Rhine River carries large positive anthropogenic La anomalies downstream from approximately Rhine‐km 450. In some cases, especially closer to the site of contamination, other LREE such as Ce and Pr are also in overabundance in the Rhine River. This leaves Sm, Nd and Eu as suitable elements for calculating La anomalies. Due to the

210

Appendix 11 continued.

possibility of small but natural Eu anomalies, the use of Nd and Sm is well suited to the dataset presented here. More recently, large anthropogenic Sm anomalies have been measured in the Rhine River complicating the calculation of REE anomalies and raising the issue of consistency. Ideally, Nd, Sm and Eu should all lie on a straight line in a semi‐logarithmic REE plot. The choice of Sm or Eu strongly depends on the presence of anomalies for each element. A close inspection of the REE patterns is instrumental in this case. Large Sm anomalies downstream of Rhine‐km 447.3 have emerged sometime between mid‐October 2010 and mid‐May 2011, after the collection of a large fraction of the samples presented here. This, coupled with a few samples

* exhibiting non‐unity EuSN/Eu SN ratios, Nd and Sm will be used for calculating the theoretical line that background values for La, Ce, Eu and Gd ideally pass through (Fig. 1).

* The y value (logR SN) of a point with x‐value (REE number) along this straight line can be written as follows, as simplified from the more general formula above adapted for two points:

* where R stands for any REE, signifies background concentrations, SN signifies normalization to

PAAS (Post Archean Average Australian Shale; McLennan, 1989). Starting with xLa = 1, xCe = 2 and so on, until xGd = 8, while y1 = logNdSN and y2 = logSmSN. Inserting the necessary x values and log concentrations of Nd and Sm into the above equation, the following can be derived:

∗ ∗ ∗ 5 3⁄2

∗ ∗ ∗ 2

∗ ∗ ∗ 3 ⁄2

211

Appendix 11 continued.

∗ ∗ ∗ 2

* For each such calculated logR SN, the related background concentration and anthropogenic concentration can be calculated using:

∗ ∗

∗

where [RPAAS] is the concentration of R in PAAS.

* Using equations above, anomalies of La, Ce, Eu and Gd have been quantified (LaSN/La SN, * * * CeSN/Ce SN, EuSN/Eu SN and GdSN/Gd SN) as well as (extrapolated) natural background concentrations (La*, Ce*, Eu* and Gd*) and anthropogenic concentrations where appropriate

(Laanthr and Gdanthr).

In Chapter III, background concentrations of La and Gd are calculated using Nd and Sm due to the influence of the LREE contamination at Rhine‐km 447.3 on Pr. In Chapter 6, for tap water samples from Berlin, the elements Pr, Nd and Sm are used in order to minimize errors due to uncertainty in single elemental concentrations. In general, the use of more REE is suggested to get a better fit for the linear curve estimating background REE, although the choice of REE may be restricted in some cases, such as the Rhine River downstream of Rhine‐km 447.3.

212

Appendix 11 continued.

References

Bau, M., 1999. Scavenging of dissolved yttrium and rare earths by precipitating iron oxyhydroxide: experimental evidence for Ce oxidation, Y‐Ho fractionation, and lanthanide tetrad effect. Geochimica et Cosmochimica Acta 63, 67‐77.

Schijf, J., Byrne, R.H., 2001. Stability constants for mono‐ and dioxalato‐complexes of Y and the REE, potentially important species in groundwaters and surface freshwaters. Geochimica et Cosmochimica Acta 65, 1037‐1046.

213