Estuarine, Coastal and Shelf Science 288 (2023) 108359

Contents lists available at ScienceDirect

Estuarine, Coastal and Shelf Science

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

Seasonal variations of dissolved rare earth elements and anthropogenic

gadolinium in the highly urbanized river basin, Busan, Korea
Ijin Lim a, Chulin Sun b, Jun-Ho Lee c, Jiwoo Kim a, Sumin Lee a, Hunsub Sim a, Hyung-Mi Cho d,
Jong-Sik Ryu a, Taejin Kim a, *
a
Division of Earth Environmental System Science, Pukyong National University, 45 Yongso-ro, Nam-gu, Busan, 48513, Republic of Korea
b
Marine Environment Monitoring Team, Korea Marine Environment Management Corporation, 22 Haeyang-ro 301, Yeongdo-gu, Busan, 49111, Republic of Korea
c
Training Ship NARA, Pukyong National University, 45 Yongso-ro, Nam-gu, Busan, 48513, Republic of Korea
d
Department of Ocean Sciences, Inha University, 100 Inha-ro, Michuhol-gu, Incheon, 22212, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Anthropogenic gadolinium (Gdanth) has been reported in rivers as a result of magnetic resonance imaging. We
Rare earth elements collected seasonal water samples from the Suyeong River Basin, to investigate seasonal variations in dissolved
Gadolinium rare earth elements (REEs) and Gdanth in highly urbanized areas. The post-Archean Australian shale-normalized
Seasonal variation
REE patterns indicate a significant influence of terrestrial source inputs during the summer. Although higher
Gdanth sources
Suyeong river basin
concentrations of dissolved REEs were measured in summer than in winter, the Gdanth concentrations exhibited
similar levels. During winter, the concentrations of Gdanth showed conservative mixing trends downstream of the
wastewater treatment plant (WWTP), whereas no clear trend was observed in summer. Moreover, the estimated
daily discharge of Gdanth accounts for approximately 11–22% of the number of MRI tests performed, when the
WWTP treats only 6% of Busan area size. Our results suggests that the elevated Gdanth level in the highly ur­
banized river basin is a major concern.

1. Introduction water-soluble Gd-based contrast agents, such as

Gd-diethylenetriaminepentaacetic acid (DTPA). Gd-DTPA and other Gd
Rare earth elements (REEs), comprising 15 lanthanide elements (La complexes (e.g., Gd-DOTA, Gd-DTPA-BMA, and Gd-BOPTA) are used in
to Lu), are an extremely coherent group in terms of their chemical magnetic resonance imaging (MRI) (Bau and Dulski, 1996a; Ku¨nne­
behavior. In aquatic environments, all REEs except cerium (Ce4+) and meyer et al., 2009). Previous studies have revealed that some Gd che­
europium (Eu2+), behave similarly because of their trivalent oxidation lates are broken down in vivo and accumulate in tissues, bones, and the
states (Elderfield and Greaves, 1982; Nozaki, 2001). The industrial de­ brain (Thakral et al., 2007; Darrah et al., 2009, 2013). Owing to the high
mand for REEs has recently increased due to the proliferation of stability of these Gd complexes, most of the Gd used for MRI is excreted
advanced technologies, including computer memory, rechargeable from the body within a few hours without being metabolized (Küm­
batteries, cell phones, liquid crystal displays, and catalytic converters merer and Helmers, 2000; Möller et al., 2002). Subsequently, it is
(Massari and Ruberti, 2013). Thus, over the past few decades, the use directly introduced into the water without being removed during
and subsequent waste of these goods has discharged anthropogenic REEs wastewater treatment (Möller et al., 2003; Inoue et al., 2022), where it
into the environment. can then be absorbed by plants and organisms in aquatic systems (Lin­
Among the REEs, gadolinium (Gd) can be used to trace anthropo­ gott et al., 2016). The occurrence of anomalously high Gd concentra­
genic sources (Bau and Dulski, 1996a; Hatje et al., 2016; Kim et al., tions in the environment was first reported in rivers in Germany (Bau
2020b; Louis et al., 2020; Wang et al., 2021; Liu et al., 2022). The major and Dulski, 1996a). A study in San Francisco Bay recently reported a
sources of anthropogenic Gd (Gdanth) in aquatic systems are marked increase in Gd over the past two decades (1993–2013) by the

Abbreviations: Gdanth, anthropogenic gadolinium; REE, rare earth element; PAAS, post-Archean Australian shale; MRI, magnetic resonance imaging; WWTP,
wastewater treatment plant; LDPE, low-density polyethylene; ICP-MS, inductively coupled plasma mass spectrometry; LREE, light rare earth element; MREE, middle
rare earth element; HREE, heavy rare earth element.
* Corresponding author.
E-mail addresses: [email protected] (I. Lim), [email protected] (T. Kim).

https://doi.org/10.1016/j.ecss.2023.108359
Received 8 May 2022; Received in revised form 19 April 2023; Accepted 28 April 2023
Available online 28 April 2023
0272-7714/© 2023 Elsevier Ltd. All rights reserved.
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

rapidly expanding the use of Gd as an MRI contrast agent and suggested

that Gd could be an emerging contaminant in the future (Hatje et al.,
2016), although the Gd concentration level (pM) was well below the
threshold of ecotoxicological effects estimated through culture experi­
ments (~0.4 μM) (González et al., 2015). Therefore, Gdanth has become a
great concern in aquatic environments, because of its elevated concen­
trations and the potential toxicity to aquatic ecosystems.
Although many studies have identified positive Gd anomalies in
urbanized rivers and estuaries (Nozaki et al., 2000; Bau et al., 2006;
Kulaksız and Bau, 2007; Lawrence, 2010; Hatje et al., 2016; Song et al.,
2017; Kim et al., 2020a, 2020b), only a few have investigated seasonal
variations (Ryu et al., 2007; Möller et al., 2014; Smith and Liu, 2018).
These studies revealed that different amounts of dissolved REEs with
positive Gd anomalies can be discharged into aquatic environments.
Therefore, in this study, we investigated seasonal variations in dissolved
REE distributions in the Suyeong River Basin, Busan, Korea. The study
region was selected base on the high discharge probability of anthro­
pogenic REEs into the aquatic environment due to the surrounding
highly urbanized areas. Furthermore, using Gd anomalies, Gdanth can be
evaluated quantitatively.

2. Materials and methods

2.1. Study area

The Suyeong River is the second longest river in Busan (the second
largest city in Korea; ~3.5 million population; 3600 people km− 2 pop­
ulation density), which is 28.6 km long with a watershed area of 199
km2 (Kim et al., 2019). The river flows south from the Hoedong Reser­
voir to Suyeong Bay, with two tributaries, Oncheon and Seokdae. The
Heodong–Seokdae industrial complexes are located in the upper region
of the river, upstream of the joining point between the main channel and
the Seokdae stream. The Suyeong wastewater treatment plant (WWTP)
treats wastewater from the highly urbanized areas (~7800 people km− 2
Fig. 1. Map showing the Suyeong River Basin and sampling locations. The
population density) in Busan and is located downstream of the Oncheon
squares (February 2020; 20W), circles (August 2020; 20S), and triangles
stream confluence (Fig. 1). Furthermore, several drainage pumping (January 2021; 21W) indicate sampling sites. The industrial complexes and the
stations discharge rainwater into the Suyeong River. The annual pre­ Suyeong wastewater treatment plant (WWTP) are also shown.
cipitations in this study period (2020 and 2021) were 1810 mm and
2280 mm, respectively, and most precipitation (71–77%) occurs be­ 2.3. Sample preparation
tween June and September (Korea Meteorological Administration, https:
//www.weather.go.kr). Water samples were pre-concentrated using the column extraction
method with Chelex-100 resin (100–200 mesh, Bio-Rad, USA) (Kim
2.2. Sampling et al., 2010). All procedures for the preconcentration and extraction of
REEs from water samples were conducted in a class-100 clean room. The
Surface water samples were collected from the Suyeong River Basin Chelex-100 resin was soaked overnight in 1 M ammonium acetate so­
during both winter and summer during low tide. Samples for winter lution (NH4Ac, Sigma Aldrich, USA), before being packed into a resin
were collected in February 2020 (20W; n = 7) and January 2021 (21W; column (1.2 g wet weight). Subsequently, the resin columns were
n = 14), whereas those for summer were collected in August 2020 (20S; pre-cleaned with 2 M HNO3 (Ultra-High Pure Grade, ODLAB, Korea)
n = 15). Surface water samples in 20W were collected from the riverside, followed by Milli-Q water and HAcO-NH4AcO buffer solution (pH 6)
whilst in 20S and 21W were collected from the center of the river using a prior to sample loading. Briefly, 3.6 M HAcO-NH4AcO buffer solution
small boat. In addition, coastal seawater samples (F1 and F2) and was prepared by mixing 100 g of 5.2 M acetic acid (99%, OPTIMA,
effluent from the Suyeong WWTP were collected in August 2020 and Fisher Chemical, USA) and 41 g of NH3 (20–22%, OPTIMA, Fisher
April 2021, respectively (Fig. 1). Temperature and salinity were Chemical, USA). The acidified aqueous samples (10 mL) were adjusted
measured in the field using a portable meter (ORION Star A320; Thermo to a pH of 5.8–6.0, using HAcO-NH4AcO buffer solution and 4% NH3
Fisher, USA). Water samples were collected using an acid-cleaned plastic (20–22%, OPTIMA, Fisher Chemical, USA). The solutions were passed
bucket, poured into 500 mL acid-cleaned low-density polyethylene through a Chelex-100 resin column at a flow rate <1.2 mL min− 1.
(LDPE) bottles (Nalgene/Thermo Fisher, USA), and after rinsing three Subsequently, a 0.05 M HAcO-NH4AcO buffer solution was used to
times, were stored in clean plastic zip-lock bags. After collection, the maintain the pH of the resin and wash out the interfering elements. The
water samples were filtered with acid-cleaned 0.2 μm pore-size mem­ REEs absorbed onto the resin were then eluted with 5.0 mL of 2 M HNO3
brane filters (Mixed Cellulose Ester, Advantec, Japan) in a class-100 and stored in acid-cleaned Teflon tubes (15 mL; Savillex, USA). Finally,
clean room installed in the laboratory and stored in separately pre­ 0.1 mL of Rh (10 μg L− 1) was added to each eluate as an internal stan­
pared 500 mL acid-cleaned LDPE bottles. The samples were subse­ dard to correct for instrumental instability. Prior to sample measure­
quently acidified to pH < 2 using HCl (Ultra-High Pure Grade, ODLAB, ment, a recovery test was performed using Tm (Kim et al., 2010), to
Korea) and stored in clean plastic zip-lock bags for analysis. ensure that extraction efficiency was 102 ± 2% (n = 4). Subsequently,
the sample measurements were performed.

2
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

2.4. Elemental analysis (0.2–24.0). This implies that seawater can enter the river more easily in
winter owing to lower river discharge rates.
The concentrations of the dissolved REEs in the eluates were deter­
mined using inductively coupled plasma mass spectrometry (ICP-MS; 3.2. Dissolved REEs concentrations
NexION 300, PerkinElmer, USA). The REE concentrations were deter­
mined using an external standard curve created by diluting the standard The dissolved REE concentrations generally varied by one to two
solution. The procedural blanks for La, Pr, Nd, Sm, Eu, Tm, Yb, and Lu orders of magnitude in 20W and 20S, but not in 21W. For example,
were 0.18, 0.07, 0.36, 0.16, 0.46, 0.02, 1.03, and 0.40 pM, respectively, dissolved Nd concentrations ranged from 17.3 to 193 pM in 20W, and
whereas other REEs were undetectable (n = 7). The detection limit (3σ) 64.6–6310 pM in 20S. In contrast, dissolved Nd concentrations
of Yb is 2.25 pM, whereas those of the other elements are <1.13 pM. measured in 21W had lower degrees of variation, ranging from 12.8 to
Through measurement of the test sample, we found that BaO acted as 21.0 pM (Table 2 and Fig. 2). In an estuarine environment, dissolved
an interfering element for Eu, which has also been reported in a previous REEs are far below the conservative mixing lines between river water
study (Wang et al., 2021). To correct this, we determined the concen­ and seawater, because of the dissolved REE losses in the low-salinity
trations of Ba isotopes in both the Ba standard solution and water region. These losses are caused by salt-induced coagulation of river-
samples, and then calculated the corrected concentrations of dissolved borne colloids and adsorption onto estuarine particles and sediments
153
Eu using the following equations: in the order of light REEs (LREEs; from La to Nd) > middle REEs (MREEs;
from Sm to Dy) > heavy REEs (HREEs; from Ho to Lu) (Sholkovitz, 1993;
I(153) = 153Eu+ + 137BaO+ Kulaksız and Bau, 2007). Conversely, a previous study in which estuary
137
BaO+ = 137Ba+ ⋅ KO(137Ba/135Ba) ⋅ (135BaO+/135Ba+) dams were present, showed the inverse correlations between dissolved
REEs and salinity via the rapid mixing of seawater and river water
which produces the true intensity of 153Eu+. Prior to sample analysis, resulting from sudden discharges of river water when the estuary dam
KO(137Ba/135Ba), the ratio of 137Ba and 135Ba oxide production rate, was was opened (Kim et al., 2020b). However, in this study, the dissolved
calculated based on signal intensities measured in the Ba standard so­ REE concentrations exhibited no significant correlation with salinity,
lutions. During sample measurements, the concentrations of 135Ba+, except for some HREEs in winter. In 20W, only two HREEs (Yb and Lu)
137 +
Ba , and I(153) were measured for each sample solution (Nakamura exhibited inverse correlations with salinity (R2 = 0.66–0.80), whereas
and Chang, 2007). the other REEs exhibited little correlation (R2 ≤ 0.01–0.46). Similarly,
To check the validity of this method, we measured a reference only four HREEs (Er, Tm, Yb, and Lu) exhibited inverse correlations with
sample (SLEW-3, National Research Council of Canada). Although salinity in 21W (R2 = 0.70–0.91). However, in 20S, the dissolved REE
certified REE values were not available, the values measured using this concentrations increased as salinity decreased, and there were no sig­
method were consistent with those reported in previous studies for all nificant correlations between dissolved REEs and salinities (R2 ≤ 0.18).
REEs (Lawrence and Kamber, 2007; Kim et al., 2020b) (Table 1). This could be caused by dissolved REE inputs from various sources.
Overall, the dissolved REE concentrations were relatively high in 20S
3. Results and discussion (total dissolved REE concentration; ΣREE = 471–32900 pM) compared
with those in winter (20W and 21W, ΣREE = 164–1390 pM) (Table 2
3.1. General parameters and Fig. 2). Considering the relatively high precipitation rates in sum­
mer, lower concentrations of dissolved REEs due to dilution are expected
Summer water temperatures (20S) ranged from 21.2 to 27.1 ◦ C in 20S. Though, in this study, generally higher dissolved REE concen­
(Table 2), whereas winter temperatures (20W and 21W) ranged from trations were measured compared with those in winter (20W and 21W).
10.4 to 14.2 ◦ C and from 10.1 to 13.3 ◦ C, respectively. Although salinity These higher concentrations were likely due to the additional input of
generally increased closer to the bay, some sampling sites exhibited dissolved REEs from stream water and drainage pumping stations,
exceptional values (e.g., 17.4 at SR4 in 20W; Table 2). These results which could also include rainwater runoff.
indicate the influence of water discharges from drainage pumping sta­ In 20S, considerably higher concentrations of dissolved REEs were
tions and streams rather than simple mixing between the river and obtained at SR4 (e.g., Nd = 6310 pM, Gd = 1200 pM, and Lu = 60.4 pM)
seawater. Furthermore, the salinities during winter (20W and 21W; (Table 2 and Fig. 2), which is near the Suyeong WWTP and Oncheon
13.2–34.3) were relatively high compared with those obtained in 20S stream confluence (Fig. 1). Therefore, these higher concentrations of
dissolved REEs could be due to effluents from the Suyeong WWTP and/
or water from the Oncheon stream. Nevertheless, these considerably
Table 1
higher concentrations of dissolved REEs were not observed in 20W and
Measured concentrations of rare earth elements (n = 6) for the certified mate­
rials (SLEW-3).
21W, and the dissolved REE concentrations from the Suyeong WWTP
effluent were considerably lower than those obtained at SR4 in the 20S
Elements SLEW-3 (pM)
(Table 2). Therefore, these results suggest that the high concentrations
(a) (b) This study of dissolved REEs at SR4 in the 20S could be affected by the Oncheon
La 56.8 ± 3.3 57.2 ± 0.6 56.8 ± 9.7 stream rather than the Suyeong WWTP.
Ce 53.9 ± 6.2 53.1 ± 1.5 48.4 ± 6.7 In 20W, the highest concentrations of dissolved REEs were observed
Pr 12.1 ± 0.6 12.3 ± 0.2 10.4 ± 1.3 at SR9, except for Lu, which was observed upstream of SR4 (Table 2).
Nd 58.4 ± 5.5 57.9 ± 0.8 49.2 ± 6.6
However, these features observed at SR9 were not observed in 21W,
Sm 46.5 ± 2.5 47.6 ± 1.9 43.5 ± 6.0
Eu 3.94 ± 1.25 4.01 ± 0.07 4.04 ± 0.88 where lower concentrations of dissolved REEs were measured (e.g., Nd
Gd 19.6 ± 1.9 19.3 ± 0.4 17.2 ± 2.1 = 16.4 pM; Gd = 56.4 pM), compared with 20W (e.g., Nd = 193 pM; Gd
Tb 2.83 ± 0.38 2.96 ± 0.19 2.62 ± 0.47 = 198 pM). This indicates sporadic inputs of dissolved REEs from an
Dy 20.6 ± 2.3 21.0 ± 0.7 18.7 ± 1.7 additional source, such as the drainage pumping station located near
Ho 5.52 ± 0.79 5.57 ± 0.06 4.84 ± 0.63
Er 16.3 ± 2.2 16.5 ± 0.5 15.6 ± 2.3
SR9. Our results demonstrate the complexity of the study region, with
Tm 2.19 ± 0.41 2.21 ± 0.46 1.78 ± 0.25 the varying inputs of dissolved REEs from various sources during
Yb 11.8 ± 3.1 11.8 ± 0.6 10.1 ± 1.7 different seasons.
Lu 1.89 ± 0.63 1.91 ± 0.05 1.79 ± 0.43

(a) Values reported from Lawrence and Kamber (2007), (n = 11).

(b) Values reported from Kim et al. (2020b), (n = 4).

3
 I. Lim et al.
Table 2
The concentrations of dissolved REEs (pM) with temperature, salinity, YbSN/NdSN ratio, Gd anomaly (Gd/Gd*), and estimated concentrations of anthropogenic Gd from anthropogenic origin (pM) in the Suyeong River in
winter (20W and 21W) and summer (20S) season.
Station Temp. (◦ C) Sal. (psu) Concentrations of dissolved REEs (pM) YbSN/NdSN Gd/Gd* Gdanth (pM)

20W La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

SR2 10.4 21.5 41.8 36.1 36 17.3 3.45 0.74 147 0.77 6.59 2.20 10.9 2.2 22.2 4.30 18 36 142
SR4 10.9 17.4 50.1 41.5 15 23.8 5.04 1.1 90.1 1.1 9.93 3.20 14.5 3.0 28.7 6.10 17 15 83.1
SR5 13.5 18.2 48.1 49.3 44 21.9 4.40 0.97 225 0.96 7.94 2.55 11.8 2.4 25.2 4.98 17 44 219
SR9 13.9 21.5 270 504 5.6 193 40.2 7.6 198 6.1 36.0 7.68 25.9 4.2 33.8 5.96 2.5 5.6 157
SR11 14.2 21.1 61.8 72.7 29 24.7 4.76 1.0 152 0.98 7.98 2.48 11.5 2.2 21.8 4.25 13 29 146
SR14 13.1 25.9 49.2 48.7 22 21.6 4.15 0.91 107 0.90 7.09 2.07 8.84 1.7 16.3 3.04 11 22 102
SR18 13.5 32.7 38.8 25.6 4.8 21.6 4.82 1.2 30.2 1.2 9.82 2.47 8.84 1.4 10.5 1.83 7.0 4.8 22.9

20S

SR1 25.5 0.2 265 385 63.9 281 58.0 15 74.1 8.1 51.2 11.3 37.1 5.7 37.6 6.70 1.9 1.5 18.0
SR2 26.3 0.2 855 1540 199 794 146 30 158 21 122 24.4 71.8 9.9 62.9 9.85 1.1 1.3 14.5
SR3 26.5 0.4 1990 3820 454 1870 332 71 355 51 277 52.8 154 20 127 20.7 0.98 1.2 13.2
SR4 27.1 0.3 6320 13700 1560 6310 1180 230 1200 170 932 178 512 66 414 60.4 0.95 1.2 27.6
SR5 26.1 1.5 251 400 55.2 233 43.2 10 91.1 6.0 38.8 8.16 26.7 4.4 32.2 7.15 2.0 2.5 49.2
SR6 25.6 1.8 874 1690 201 810 146 30 190 21 126 24.7 70.7 10 70.5 11.9 1.2 1.5 46.4
SR7 25.6 3.1 836 1500 181 728 126 27 258 19 104 21.5 63.6 9.3 64.0 10.8 1.3 2.3 130
SR9 25.7 4.2 587 1060 126 505 88.6 19 363 14 80.5 16.3 52.1 7.6 56.2 10.2 1.6 4.5 270
SR11 26.0 4.0 485 860 104 421 75.5 16 226 11 65.7 13.7 44.1 6.4 49.7 9.49 1.7 3.4 148
SR13 26.2 4.5 594 1050 128 521 94.9 18 225 14 77.0 15.8 50.9 7.7 55.4 10.4 1.5 2.8 131
4

SR14 25.9 6.4 654 1170 141 547 104 24 239 15 87.6 17.8 55.3 8.1 57.9 11.3 1.5 2.6 135
SR15 25.1 7.6 645 1150 140 551 97.4 18 223 14 83.6 17.0 54.1 8.3 54.4 9.79 1.4 2.6 59.8
SR16 24.4 11.8 311 482 57.6 234 40.5 8.8 157 6.2 38.0 8.43 28.1 4.6 35.8 7.46 2.2 4.3 180
SR17 24.8 9.2 238 370 44.8 188 34.1 6.9 157 4.9 29.4 7.05 26.4 4.0 34.7 7.30 2.7 5.4 123
SR19 21.2 24.0 145 115 16.4 64.6 10.3 1.9 71.2 1.7 11.7 3.13 11.3 1.7 13.4 3.11 3.0 7.4 60.0
F1 18.4 31.2 108 61.9 16.7 69.3 13.0 3.2 21.2 3.1 18.6 4.43 14.6 2.3 13.9 2.44 2.9 1.3 2.66
F2 17.6 31.6 110 55.9 16.4 67.3 12.2 3.2 20.9 2.8 16.2 4.41 12.7 2.0 11.6 2.50 2.5 1.4 3.78

21W

SR4 10.1 13.2 23.0 17.6 3.52 14.0 3.35 2.0 62.9 0.78 6.72 2.44 13.7 2.6 24.8 5.87 26 15 58.2
SR5 10.2 14.4 18.8 16.3 3.19 12.8 2.87 2.3 67.1 0.71 7.75 2.48 12.3 2.5 23.1 5.31 26 18 62.8
SR6 12.0 15.6 26.2 26.4 3.78 13.6 2.95 1.6 147 0.76 6.49 2.41 9.77 2.0 18.8 4.42 20 38 142

Estuarine, Coastal and Shelf Science 288 (2023) 108359

SR7 12.6 18.3 31.4 32.6 4.04 17.0 3.66 1.8 143 0.69 6.26 1.98 9.05 1.8 15.8 3.70 13 38 138
SR8 12.6 16.9 29.0 30.5 3.73 14.0 2.39 1.8 157 0.74 6.44 2.16 10.1 1.8 20.0 4.28 21 43 152
SR9 13.0 28.7 37.6 33.6 4.62 16.4 2.71 1.4 56.4 0.52 4.90 1.62 6.63 0.99 9.71 1.97 8.5 20 53.1
SR10 13.0 21.6 35.6 37.0 4.97 19.2 2.79 0.94 125 0.77 6.12 2.13 8.10 1.6 13.3 3.01 10 32 120
SR11 13.1 22.1 38.5 39.8 4.39 18.5 3.19 1.4 120 0.74 6.12 2.06 8.13 1.3 14.6 2.89 11 31 115
SR12 13.3 22.7 38.7 40.9 4.57 18.2 3.51 1.1 115 0.71 4.99 1.88 8.20 1.5 12.8 2.92 10 30 110
SR13 13.2 23.5 38.5 41.4 5.16 18.9 4.78 1.5 106 0.81 6.63 1.96 7.79 1.4 12.5 2.70 9.5 23 101
SR14 13.2 24.9 38.1 38.8 4.64 16.2 2.31 0.90 86.6 0.71 6.12 1.48 7.58 1.3 12.5 2.33 11 24 82.4
SR16 13.2 25.4 37.8 39.3 4.63 16.8 3.04 1.7 87.4 0.70 6.29 1.76 6.89 1.0 11.8 2.83 10 24 83.2
SR17 12.8 26.1 34.1 33.5 4.99 17.9 2.59 0.63 71.5 0.71 5.81 1.73 6.54 1.3 9.67 2.69 7.8 20 67.3
SR19 12.5 34.3 32.4 21.7 5.02 21.0 4.29 1.1 9.99 1.2 7.78 2.16 7.44 0.96 6.90 1.42 4.7 1.7 3.01
WWTP 14.0 25.9 4.33 20.1 3.71 0.86 560 1.4 14.2 5.59 27.4 6.0 63.5 11.3 46 84 552
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

sedimentary rocks (Pedreira et al., 2018; Kim et al., 2020a). Therefore,

this temporal difference between 20W and 21W at SR9 implies an
additional terrestrial source input in this area, which occasionally
discharges.
In contrast to winter (20W and 21W), relatively flat patterns were
observed in summer (20S). The upper sampling sites (SR1–SR4), YbSN/
NdSN ratios were calculated as 1.2 ± 0.4 (n = 4) (Fig. 3c). Interestingly,
at SR3 (ΣREE = 9600 pM) and SR4 (ΣREE = 32900 pM), higher REE
concentrations with lower YbSN/NdSN ratios (≤0.98) were observed
compared to SR1 (ΣREE = 1300 pM; YbSN/NdSN = 1.9) and SR2 (ΣREE
= 4050 pM; YbSN/NdSN = 1.1), although these sampling sites were
located midstream near the Oncheon stream confluence (Fig. 1). Thus,
these enriched and flat REE patterns may indicate the input of the
Oncheon stream. At sampling sites between SR5 and SR17, where HREE
enrichment was observed in 20W and 21W (YbSN/NdSN = 13 ± 6, n =
16), relatively flat patterns or slightly enriched HREEs were observed in
20S (YbSN/NdSN = 1.7 ± 0.4, n = 10) (Fig. 3d). Although the Suyeong
WWTP may discharge wastewater consistently, the patterns obtained in
20S appeared like those observed at SR9 in 20W (YbSN/NdSN = 2.5)
(Fig. 3a). These different patterns compared to winter suggest that the
sampling sites in 20S are significantly influenced by inputs from
terrestrial sources, which could conceal the signal from the WWTP. At
SR19 in 20S, which is the outermost sampling site in the Suyeong River
and had a salinity of 24.0, the PAAS-normalized pattern indicated
mixing between river water and coastal seawater (F1 and F2), with
seawater salinities measured at 31.2–31.6 (Fig. 3e and Table 2).

3.4. Positive Gd anomalies in the Suyeong River Basin

In this study, positive Gd anomalies were consistently observed at all

sampling sites during all seasons (Fig. 3). The Gd anomaly was calcu­
lated using the following equation:

Gd anomaly = GdSN/GdSN*

GdSN/GdSN* = GdSN/(0.33SmSN + 0.67TbSN)

where * denotes the geogenic Gd source interpolated between neigh­

boring elements (Sm and Tb) and SN represents the PAAS normalization
Fig. 2. Dissolved Nd, Gd, and Lu concentrations at each sampling site. Nd, Gd,
(Knappe et al., 2005). In natural river waters unaffected by anthropo­
and Lu are representative of LREE, MREE, and HREE, respectively.
genic REEs, negative Ce anomalies and either negative or positive Eu
anomalies are commonly observed (Lewis et al., 1997; van Middles­
3.3. Shale-normalized REE patterns worth and Wood, 1998; Kidder et al., 2003; Kulaksız and Bau, 2011b;
Song et al., 2017). Furthermore, a small positive Gd anomaly (<1.4) is
The post-Archean average Australian shale (PAAS)-normalized REE common because of the “tetrad effect,” which results from the half-filled
patterns (McLennan, 1989) are presented in Fig. 3 to eliminate the 4f electron in its orbital shell, irrespective of the normalization problem
Oddo–Harkins effect in natural abundances, allowing easy comparison (e.g., Bau and Dulski, 1996b; Alibo and Nozaki, 1999). Therefore, we
with other REE data (Migaszewski and Gałuszka, 2015). considered a value of 1.4 as the baseline for anthropogenic Gd anomalies
During winter (20W and 21W), all sampling sites, except SR9 in (Fig. 4).
20W, exhibited HREE enrichment relative to LREE (Fig. 3a and b). YbSN/ Throughout the Suyeong River Basin and across the seasons, we
NdSN ratios, a commonly used representative for HREESN/LREESN, in observed positive Gd anomalies, and most of them showed distinct
20W and 21W ranged from 7.0 to 18 (average: 14 ± 4, n = 6) and from anthropogenic Gd anomalies. In this study, the calculated Gd anomalies
4.7 to 26 (average: 14 ± 7, n = 14), respectively (Table 2). This pattern in 20W, 21W, and 20S ranged from 4.8 to 50 (average: 22 ± 14, n = 7),
could reflect the preferential scavenging of LREEs from river water in the 1.7 to 43 (average: 26 ± 10, n = 14), and 1.2 to 7.4 (average: 3.0 ± 1.7,
low-salinity region via coagulation, aggregation, and sedimentation of n = 15), respectively, whereas that of the WWTP effluent exhibited a
the river colloids as freshwater mixes with seawater (Sholkovitz, 1993; considerably higher value (Gd/Gd* = 84) (Table 2 and Fig. 4). The Gd
Kulaksız and Bau, 2007). Interestingly, a similar pattern was observed in anomaly in the WWTP effluent obtained in this study was significantly
the effluent of the Suyeong WWTP, which also exhibited HREE enrich­ higher than that reported in Montreal (Gd/Gd* = 3.2), but within the
ment (YbSN/NdSN = 46) with substantially enriched Gd anomaly range of those reported in Seoul (57–198), Wuhan (30–158), and South
(Table 2 and Fig. 3b). This could be due to the oxidative removal of Fe East Queensland (9–99) (Lawrence et al., 2009; Song et al., 2017; Wang
during wastewater treatment, which results in preferential LREE et al., 2021; Dang et al., 2022). These anomalies can be primarily
removal (Leybourne and Johannesson, 2008; Verplanck et al., 2010; attributed to the large inputs of refractory Gd-based contrast agents,
Song et al., 2017). such as Gd-DTPA, from MRI tests (Bau and Dulski, 1996a; Möller et al.,
In contrast, a relatively flat pattern (YbSN/NdSN = 2.5) was obtained 2000; Bau et al., 2006).
at SR9 in 20W, whereas this feature was not observed in the next year During winter (20W and 21W), Gd anomalies (Gd/Gd* = 22 ± 14
(21W) (Fig. 3a). This flat pattern likely reflects the influence of terres­ and 26 ± 10) were conspicuously higher than those in 20S (Gd/Gd* =
trial source inputs from the chemical weathering of widespread 3.0 ± 1.7) (Table 2 and Fig. 4). These Gd anomalies in winter are

5
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

Fig. 3. Post-Archean Australian shale (PAAS)-normalized patterns of dissolved REEs. (a) February 2020 (20W), (b) January 2021 (21W), (c)–(e) August 2020 (20S).

comparable to those previously reported for the Rhine River (Gd/Gd* = it does not consider the natural Gd anomaly, which is 1.16 ± 0.28 in
22.5 ± 29.3) (Kulaksız and Bau, 2011a) (Table 3). At SR9 in 20W, where most river systems (Dang et al., 2022). Accordingly, Gdanth was calcu­
a flat PAAS-normalized REE pattern was observed (Fig. 3a), the Gd lated as follows:
anomaly was lower (Gd/Gd* = 5.6) than those of adjacent sampling ∗
Gdanth = Gdmeas – Friver
natural × Gd
sites (Gd/Gd* = 44 at SR5, and 29 at SR11), although the dissolved Gd
concentration (198 pM) was similar (Gd = 225 pM at SR5, and 152 pM
where Gdmeas indicates the measured dissolved Gd concentration, and *
at SR11; Table 2). As previously mentioned, high concentrations of
denotes the geogenic Gd source interpolated between neighboring ele­
dissolved REEs (ΣREE = 1390 pM) are released near SR9. Considering
ments (Sm and Tb). Friver
natural indicates the correction factor to consider the
that Gd anomalies were calculated based on interpolation from neigh­
natural Gd anomaly (Friver
natural = 1.16) (de Baar et al., 2018).
boring elements (Sm and Tb), a high background concentration of Gd
Gdanth concentrations in 20W, 21W, and 20S ranged from 22.9 to 219
(Gd*) could be the cause of the relatively low Gd anomaly.
pM (average: 125 ± 57 pM, n = 7), 3.07–152 pM (average: 92.1 ± 39.9
This trend was also observed in 20S, where enriched REEs with
pM, n = 14), and 13.2–270 pM (average: 93.8 ± 71.3 pM, n = 15),
relatively flat PAAS-normalized REE patterns were obtained (Fig. 3c, 3d,
respectively (Table 2). Even though the measured Gd concentrations
and 3e). In 20S, the dissolved Gd concentrations ranged from 71.2 to
were varied, the average Gdanth concentrations in the Suyeong River
1200 pM. Although this range is similar to previously reported data from
Basin were similar in all the seasons. These average Gdanth concentra­
the Rhine (75–1200 pM) and Nakdong (21.2–1660 pM) rivers, Gd
tions were higher than those in the Han River, Korea (33.5 ± 18.4 pM, n
anomalies in 20S (Gd/Gd* = 3.0 ± 1.7) were considerably lower than
= 24), Allegheny River (10.1 pM), Monongahela River (12.9 pM), and
those reported in the Rhine River (Gd/Gd* = 22.5 ± 29.3, n = 14;
Ohio River (38.4 pM) in the United States (Bau et al., 2006; Kim et al.,
Kulaksız and Bau, 2011a), but slightly higher than those reported in the
2020a), where the watershed areas are several orders of magnitude
Nakdong River (Gd/Gd* = 1.6 ± 0.3, n = 17; Kim et al., 2020b)
higher (>19,000 km2) than that of the Suyeong River (199 km2).
(Table 3), where a substantial release of all REEs was suggested for low
The spatial distributions of Gdanth in the Suyeong River Basin are
Gd anomalies (Kim et al., 2020b). Thus, the relatively low Gd anomalies
shown in Fig. 5. During winter (20W and 21W), high concentrations of
in 20S can be explained by the high background concentrations of Gd
Gdanth were observed at sites close to the WWTP (SR5–SR8). This result
(Gd*) caused by additional REE inputs from terrestrial sources.
may be related to the Gdanth discharge from the Suyeong WWTP. Sub­
sequently, the Gdanth concentrations gradually decreased downstream.
Negative correlations (R2 = 0.97) between Gdanth and salinity were
3.5. Impact of anthropogenic Gd on Suyeong River Basin
observed downstream of the WWTP, indicating that Gdanth from the
Suyeong WWTP may be conservatively mixed in winter (Fig. 6). The
To investigate the influence of Gdanth on the Suyeong River Basin,
extrapolated Gdanth concentrations were 433 pM for 20W and 286 pM
Gdanth concentrations were calculated. Based on Gd anomalies, Gdanth is
for 21W, respectively, suggesting that Gdanth discharges from the WWTP
commonly defined as the difference between the measured Gd concen­
to the Suyeong River Basin could be temporally varied. By contrast, in
trations (Gdmeas) and background Gd values (Gd*). However, a recent
summer (20S), high concentrations of Gdanth were observed at SR9 and
study mentioned that this direct calculation could overestimate Gdanth as

6
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

Table 3
Reported dissolved Gd concentration and Gd anomaly from other relevant
studies.
Study area Gd Gd anomaly (Gd/ Reference
concentration Gd*)
(pM)

Tokyo Bay 8.4–15.4 1.5–2.1 n Nozaki et al.

(average: 1.8 = (2000)
± 0.2) 5
San Francisco Bay 14.1–171 1.6–3.9 n Hatje et al.
(average: 2.3 = (2016)
± 0.8) 8
Han River 45.9–209 1.6–8.2 n Song et al.
(average: 4.3 = (2017)
± 2.5) 6
Han River 39.9–233 1.2–2.3 n Kim et al.
(average: 1.6 = (2020a)
± 0.3) 26
Nakdong River 21.2–1660 1.3–2.3 n Kim et al.
(average: 1.6 = (2020b)
± 0.3) 17
Tone, Tama, and Ara 10.3–155 1.6–6.2 n Nozaki et al.
River (average: 3.3 = (2000)
± 1.4) 12
Rhine River 75–1200 0.98–110 n Kulaksız and
(average: 22.5 = Bau (2011a)
± 29.3) 14
Havel River 3140 644 n Kulaksız and
= Bau (2011b)
Fig. 4. Gd anomaly values as the function of salinities. The open symbols 1
indicate the samples obtained upstream of the Suyeong WWTP in each season, Danube River 48.6 2.3 n Kulaksız and
while filled symbols indicate samples obtained downstream of the WWTP. The = Bau (2011b)
1
green line indicates the baseline for anthropogenic Gd anomaly value (1.4). The
River Thames 28 2.5 n Kulaksız and
dashed black line corresponds to the average Gd anomaly value (1.4) at the
= Bau (2011b)
coastal sea (F1 and F2) obtained in this study. The solid black line shows the Gd 1
anomaly value (84) in the effluent of the Suyeong WWTP. (For interpretation of Elbe, Ems, and Weser 94.8–154 3.6–5.5 n Kulaksız and
the references to colour in this figure legend, the reader is referred to the Web River (average: 4.8 = Bau (2007)
version of this article.) ± 0.9) 3
Wupper, Spree, and 43.1–675 12.4–126 n Bau and
Havel River (average: 56.1 Dulski
SR16, which were away from the WWTP (Fig. 5), and no clear mixing
=
± 49.9) 3 (1996a)
trend was observed (Fig. 6). Thus, further research is required to Suyeong 20W 30.2–225 4.8–44 n This study
determine their corresponding sources. River (Feb. (average: 22 =
To date, the influence of Gdanth on rivers has been evaluated based on 2020) ± 14) 7
Suyeong 20S 71.2–1200 1.2–7.4 n This study
the river discharge rate and Gdanth concentration (Song et al., 2017; Kim
River (Aug. (average: 3.0 =
et al., 2020a). Unfortunately, no data have been reported for either the 2020) ± 1.7) 15
annual discharge rate or seasonal variations during our sampling pe­ Suyeong 21W 9.99–157 1.7–43 n This study
riods. Therefore, the influence of Gdanth on the Suyeong River Basin was River (Jan. (average: 26 =
evaluated by calculating the daily Gdanth discharge based on the daily 2021) ± 10) 14

treatment capacity of the WWTP. Gdanth concentrations (286–552 pM)

were obtained not from the effluent sample and by extrapolation from 4. Conclusions
the linear relationships between Gdanth and salinity in winter (Fig. 6).
The results are within the same range as those previously reported in We investigated the seasonal variations in dissolved REE concen­
Seoul (266–1280 pM), Wuhan (148–558 pM), and South East Queens­ trations along the Suyeong River Basin, which is surrounded by highly
land (116–1770 pM), respectively (Lawrence et al., 2009; Song et al., urbanized areas. The dissolved REE concentrations generally exhibited
2017; Wang et al., 2021). The daily treatment capacity of the Suyeong no significant correlations with the salinity. Exceptional concentration
WWTP has been reported to be approximately 452,000 m3d-1 (Busan levels indicate the complexity of this study region, with inputs of dis­
Environmental Corporation, https://beco.or.kr). Therefore, the daily solved REEs from various sources. We observed higher concentrations of
Gdanth discharge from the WWTP was estimated as 0.13–0.25 mol d− 1 (= dissolved REEs during the summer (ΣREE = 471–32900 pM) than in the
0.007–0.014 ton y− 1). Assuming a dosage rate of 1.1 g of Gd per adult winter (ΣREE = 164–1390 pM). The flat and enriched PAAS-normalized
patient (Kümmerer and Helmers, 2000), the discharge of Gdanth to the REE patterns during summer suggest that terrestrial source inputs had a
Suyeong River Basin (0.007–0.014 ton y− 1) corresponds to 6800–13,000 considerable influence on the dissolved REE distributions during this
MRI tests y− 1. This estimation accounts for approximately 11–22% of season. Positive Gd anomalies were observed at all sampling sites during
the number of total MRI tests (60,600 tests y− 1) in Busan (Korean Sta­ each season, indicating that anthropogenic Gd is ubiquitous in the
tistical Information Service, https://kosis.kr), even though the Suyeong Suyeong River Basin. Although the Gd anomaly values show different
WWTP treats only 6% (43 km2) of Busan area (770 km2). Owing to the trends between winter and summer, the calculated Gdanth exhibited
limited data, the Gdanth discharge can vary temporally. Thus, the similar concentration levels. The Gdanth discharge from the WWTP was
calculated Gdanth discharge was only a rough estimate. Nonetheless, our estimated as 0.13–0.25 mol d− 1, which accounts for about 11–22% of
estimation suggests that the elevated Gdanth levels in the highly urban­ the number of MRI tests, when WWTP treats only 6% of Busan area size.
ized river basin are a major concern. Therefore, our results suggests that the elevated Gdanth levels in the
highly urbanized river basin are a major concern. Thus, a long-term

7
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

original draft, Conceptualization. Taejin Kim: Writing – original draft,

Supervision, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence
the work reported in this paper.

Data availability

Data will be made available on request.

Acknowledgments

This research was supported by the National Research Foundation of

Korea [NRF-2019R1G1A1099394 and NRF-2020R1C1C1008964]. We
are grateful to the crews of Training Ship NARA, Pukyong National
University, for their assistance in collecting water samples from the
Suyeong River Basin.

Fig. 5. Spatial distributions of Gdanth for each sampling season. References

Alibo, D.S., Nozaki, Y., 1999. Rare earth elements in seawater: particle association, shale-
normalization, and Ce oxidation. Geochem. Cosmochim. Acta 63, 363–372. https://
doi.org/10.1016/S0016-7037(98)00279-8.
Bau, M., Dulski, P., 1996a. Anthropogenic origin of positive gadolinium anomalies in
river waters. Earth Planet Sci. Lett. 143, 245–255. https://doi.org/10.1016/0012-
821x(96)00127-6.
Bau, M., Dulski, P., 1996b. Distribution of yttrium and rare-earth elements in the penge
and Kuruman iron-formations, Transvaal Supergroup, South Africa. Precambrian
Res. 79, 37–55. https://doi.org/10.1016/0301-9268(95)00087-9.
Bau, M., Knappe, A., Dulski, P., 2006. Anthropogenic gadolinium as a micropollutant in
river waters in Pennsylvania and in Lake Erie, northeastern United States. Chem.
Erde 66, 143–152. https://doi.org/10.1016/j.chemer.2006.01.002.
Dang, D.H., Wang, W., Sikma, A., Chatzis, A., Mucci, A., 2022. The contrasting estuarine
geochemistry of rare earth elements between ice-covered and ice-free conditions.
Geochem. Cosmochim. Acta 317, 488–506. https://doi.org/10.1016/j.
gca.2021.10.025.
Darrah, T.H., Campbell, M.E., Prustman-Pfeiffer, J.J., Poreda, R.J., Hannigan, R.E., 2013.
Trace element composition of modern human bone. In: Medical Geochemistry.
Springer, pp. 167–191. https://doi.org/10.1007/978-94-007-4372-4_10.
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. https://doi.org/10.1039/
b905145g.
de Baar, H.J., Bruland, K.W., Schijf, J., van Heuven, S.M., Behrens, M.K., 2018. Low
cerium among the dissolved rare earth elements in the central North Pacific Ocean.
Geochem. Cosmochim. Acta 236, 5–40. https://doi.org/10.1016/j.gca.2018.03.003.
Elderfield, H., Greaves, M.J., 1982. The rare earth elements in seawater. Nature 296,
214–219.
González, V., Vignati, D.A.L., Pons, M.N., Montarges-Pelletier, E., Bojic, C.,
Giamberini, L., 2015. Lanthanide ecotoxicity: first attempt to measure
environmental risk for aquatic organisms. Environ. Pollut. 199, 139–147. https://
doi.org/10.1016/J.ENVPOL.2015.01.020.
Hatje, V., Bruland, K.W., Flegal, A.R., 2016. Increases in anthropogenic gadolinium
anomalies and rare earth element concentrations in San Francisco Bay over a 20 year
Fig. 6. Gdanth concentrations as the function of salinities. The open symbols
record. Environ. Sci. Technol. 50, 4159–4168. https://doi.org/10.1021/acs.
indicate the samples obtained upstream of the Suyeong WWTP in each season, est.5b04322.
while filled symbols indicate samples obtained downstream of the WWTP. The Inoue, K., Fukushi, M., Sahoo, S.K., Veerasamy, N., Furukawa, A., Soyama, S., Sakata, A.,
dashed black line corresponds to the average Gdanth concentration (3.22 pM) at Isoda, R., Taguchi, Y., Hosokawa, S., Sagara, H., Natarajan, T., 2022. Measurements
the coastal sea (F1 and F2) obtained in this study. The solid black line indicates and future projections of Gd-based contrast agents for MRI exams in wastewater
Gdanth concentration (552 pM) in the effluent of the Suyeong WWTP. treatment plants in the Tokyo metropolitan area. Mar. Pollut. Bull. 174, 113259
https://doi.org/10.1016/j.marpolbul.2021.113259.
Kidder, D.L., Krishnaswamy, R., Mapes, R.H., 2003. Elemental mobility in phosphatic
monitoring study on both river waters and WWTP effluents is required to shales during concretion growth and implications for provenance analysis. Chem.
Geol. 198, 335–353. https://doi.org/10.1016/S0009-2541(03)00036-6.
further solidify this conclusion. Kim, I., Kim, S., Kim, G., 2010. Analytical artifacts associated with the chelating resin
extraction of dissolved rare earth elements in natural water samples. Aquat.
CRediT authorship contribution statement Geochem. 16, 611–620. https://doi.org/10.1007/s10498-010-9100-5.
Kim, I., Kim, S.H., Kim, G., 2020a. Anthropogenic gadolinium in lakes and rivers near
metrocities in Korea. Environ. Sci. Process. Impacts 22, 144–151. https://doi.org/
Ijin Lim: Writing – original draft, Visualization, Investigation, 10.1039/c9em00304e.
Formal analysis. Chulin Sun: Resources, Formal analysis. Jun-Ho Lee: Kim, S., Roh, K.J., Kim, D.M., 2019. Behavior characteristics of nonylphenol in the
downstream of river in Busan, Korea. Korean J. Fish. Aquat. Sci. 52, 74–80. https://
Resources, Investigation. Jiwoo Kim: Investigation, Formal analysis.
doi.org/10.1016/J.JALLCOM.2006.08.002.
Sumin Lee: Investigation. Hunsub Sim: Investigation. Hyung-Mi Cho: Kim, T., Kim, H., Kim, G., 2020b. Tracing river water versus wastewater sources of trace
Writing – original draft, Conceptualization. Jong-Sik Ryu: Writing – elements using rare earth elements in the Nakdong River estuarine waters. Mar.
Pollut. Bull. 160, 111589 https://doi.org/10.1016/j.marpolbul.2020.111589.

8
I. Lim et al. Estuarine, Coastal and Shelf Science 288 (2023) 108359

Knappe, A., Möller, P., Dulski, P., Pekdeger, A., 2005. Positive gadolinium anomaly in Möller, P., Dulski, P., Bau, M., Knappe, A., Pekdeger, A., Sommer-Von Jarmersted, C.,
surface water and ground water of the urban area Berlin, Germany. Geochemistry 2000. Anthropogenic gadolinium as a conservative tracer in hydrology. J. Geochem.
65, 167–189. https://doi.org/10.1016/j.chemer.2004.08.004. Explor. 69–70, 409–414. https://doi.org/10.1016/S0375-6742(00)00083-2.
Kulaksız, S., Bau, M., 2011a. Rare earth elements in the Rhine River, Germany: first case Möller, P., Knappe, A., Dulski, P., 2014. Seasonal variations of rare earths and yttrium
of anthropogenic lanthanum as a dissolved microcontaminant in the hydrosphere. distribution in the lowland Havel River, Germany, by agricultural fertilization and
Environ. Int. 37, 973–979. https://doi.org/10.1016/j.envint.2011.02.018. effluents of sewage treatment plants. Appl. Geochem. 41, 62–72. https://doi.org/
Kulaksız, S., Bau, M., 2011b. Anthropogenic gadolinium as a microcontaminant in tap 10.1016/j.apgeochem.2013.11.011.
water used as drinking water in urban areas and megacities. Appl. Geochem. 26, Möller, P., Morteani, G., Dulski, P., 2003. Anomalous gadolinium, cerium, and yttrium
1877–1885. https://doi.org/10.1016/j.apgeochem.2011.06.011. contents in the adige and isarco river waters and in the water of their tributaries
Kulaksız, S., Bau, M., 2007. Contrasting behaviour of anthropogenic gadolinium and (Provinces Trento and Bolzano/Bozen, NE Italy). Acta Hydrochim. Hydrobiol. 31,
natural rare earth elements in estuaries and the gadolinium input into the North Sea. 225–239. https://doi.org/10.1002/AHEH.200300492.
Earth Planet Sci. Lett. 260, 361–371. https://doi.org/10.1016/j.epsl.2007.06.016. Möller, P., Paces, T., Dulski, P., Morteani, G., 2002. Anthropogenic Gd in surface water,
Kümmerer, K., Helmers, E., 2000. Hospital effluents as a source of gadolinium in the drainage system, and the water supply of the city of Prague, Czech Republic.
aquatic environment. Environ. Sci. Technol. 34, 573–577. https://doi.org/10.1021/ Environ. Sci. Technol. 36, 2387–2394. https://doi.org/10.1021/ES010235Q.
es990633h. Nakamura, K., Chang, Q., 2007. Precise determination of ultra-low (sub-ng g-1) level rare
Künnemeyer, J., Terborg, L., Meermann, B., Brauckmann, C., Möller, I., Scheffer, A., earth elements in ultramafic rocks by quadrupole ICP-MS. Geostand. Geoanal. Res.
Karst, U., 2009. Speciation analysis of gadolinium chelates in hospital effluents and 31, 185–197. https://doi.org/10.1111/j.1751-908X.2007.00859.x.
wastewater treatment plant sewage by a novel HILIC/ICP-MS method. Environ. Sci. Nozaki, Y., 2001. Rare earth elements and their isotopes in the ocean. Encycl. Ocean Sci.
Technol. 43, 2884–2890. https://doi.org/10.1021/es901406g. 4, 2354–2366. https://doi.org/10.1006/rwos.2001.0284.
Lawrence, M.G., 2010. Detection of anthropogenic gadolinium in the Brisbane river Nozaki, Y., Lerche, D., Alibo, D.S., Tsutsumi, M., 2000. Dissolved indium and rare earth
plume in Moreton bay, Queensland, Australia. Mar. Pollut. Bull. 60, 1113–1116. elements in three Japanese rivers and Tokyo Bay: evidence for anthropogenic Gd and
https://doi.org/10.1016/j.marpolbul.2010.03.027. in. Geochem. Cosmochim. Acta 64, 3975–3982. https://doi.org/10.1016/S0016-
Lawrence, M.G., Kamber, B.S., 2007. Rare earth element concentrations in the natural 7037(00)00472-5.
water reference materials (NRCC) NASS-5, CASS-4 and SLEW-3. Geostand. Geoanal. Pedreira, R.M., Pahnke, K., Böning, P., Hatje, V., 2018. Tracking hospital effluent-
Res. 31, 95–103. https://doi.org/10.1111/j.1751-908X.2007.00850.x. derived gadolinium in Atlantic coastal waters off Brazil. Water Res. 145, 62–72.
Lawrence, M.G., Ort, C., Keller, J., 2009. Detection of anthropogenic gadolinium in https://doi.org/10.1016/j.watres.2018.08.005.
treated wastewater in South East Queensland, Australia. Water Res. 43, 3534–3540. Ryu, J.S., Lee, K.S., Lee, S.G., Lee, D., Chang, H.W., 2007. Seasonal and spatial variations
https://doi.org/10.1016/j.watres.2009.04.033. of rare earth elements in rainwaters, river waters and total suspended particles in air
Lewis, A.J., Palmer, M.R., Sturchio, N.C., Kemp, A.J., 1997. The rare earth element in South Korea. J. Alloys Compd. 437, 344–350. https://doi.org/10.1016/J.
geochemistry of acid-sulphate and acid-sulphate-chloride geothermal systems from JALLCOM.2006.08.002.
Yellowstone National Park, Wyoming, USA. Geochem. Cosmochim. Acta 61, Sholkovitz, E.R., 1993. The geochemistry of rare earth elements in the Amazon River
695–706. https://doi.org/10.1016/S0016-7037(96)00384-5. estuary. Geochem. Cosmochim. Acta 57, 2181–2190. https://doi.org/10.1016/
Leybourne, M.I., Johannesson, K.H., 2008. Rare earth elements (REE) and yttrium in 0016-7037(93)90559-F.
stream waters, stream sediments, and Fe-Mn oxyhydroxides: fractionation, Smith, C., Liu, X.M., 2018. Spatial and temporal distribution of rare earth elements in the
speciation, and controls over REE + Y patterns in the surface environment. Neuse River, North Carolina. Chem. Geol. 488, 34–43. https://doi.org/10.1016/j.
Geochem. Cosmochim. Acta 72, 5962–5983. https://doi.org/10.1016/j. chemgeo.2018.04.003.
gca.2008.09.022. Song, H., Shin, W.J., Ryu, J.S., Shin, H.S., Chung, H., Lee, K.S., 2017. Anthropogenic rare
Lingott, J., Lindner, U., Telgmann, L., Esteban-Fernández, D., Jakubowski, N., Panne, U., earth elements and their spatial distributions in the Han River, South Korea.
2016. Gadolinium-uptake by aquatic and terrestrial organisms-distribution Chemosphere 172, 155–165. https://doi.org/10.1016/j.chemosphere.2016.12.135.
determined by laser ablation inductively coupled plasma mass spectrometry. Thakral, C., Alhariri, J., Abraham, J.L., 2007. Long-term retention of gadolinium in
Environ. Sci. Process. Impacts 18, 200–207. https://doi.org/10.1039/C5EM00533G. tissues from nephrogenic systemic fibrosis patient after multiple gadolinium-
Liu, Y., Wu, Q., Jia, H., Wang, Z., Gao, S., Zeng, J., 2022. Anthropogenic rare earth enhanced MRI scans: case report and implications. Contrast Media Mol. Imaging 2,
elements in urban lakes: their spatial distributions and tracing application. 199–205. https://doi.org/10.1002/CMMI.146.
Chemosphere 300, 134534. https://doi.org/10.1016/j.chemosphere.2022.134534. van Middlesworth, P.E., Wood, S.A., 1998. The aqueous geochemistry of the rare earth
Louis, P., Messaoudene, A., Jrad, H., Abdoul-Hamid, B.A., Vignati, D.A., Pons, M.N., elements and yttrium. Part 7. REE, Th and U contents in thermal springs associated
2020. Understanding Rare Earth Elements concentrations, anomalies and fluxes at with the Idaho batholith. Appl. Geochem. 13, 861–884. https://doi.org/10.1016/
the river basin scale: the Moselle River (France) as a case study. Sci. Total Environ. S0883-2927(98)00019-5.
742, 140619 https://doi.org/10.1016/j.scitotenv.2020.140619. Verplanck, P.L., Furlong, E.T., Gray, J.L., Phillips, P.J., Wolf, R.E., Esposito, K., 2010.
Massari, S., Ruberti, M., 2013. Rare earth elements as critical raw materials: focus on Evaluating the behavior of gadolinium and other rare earth elements through large
international markets and future strategies. Res. Pol. 38, 36–43. https://doi.org/ metropolitan sewage treatment plants. Environ. Sci. Technol. 44, 3876–3882.
10.1016/j.resourpol.2012.07.001. https://doi.org/10.1021/es903888t.
McLennan, S.M., 1989. Rare earth elements in sedimentary rocks; influence of Wang, T., Wu, Q., Wang, Z., Dai, G., Jia, H., Gao, S., 2021. Anthropogenic gadolinium
provenance and sedimentary processes. Rev. Mineral. Geochem. 21, 169–200. accumulation and rare earth element anomalies of river water from the middle reach
https://doi.org/10.1515/9781501509032-010. of Yangtze River Basin, China. ACS Earth Space Chem. 5, 3130–3139. https://doi.
Migaszewski, Z.M., Gałuszka, A., 2015. The characteristics, occurrence, and geochemical org/10.1021/acsearthspacechem.1c00238.
behavior of rare earth elements in the environment: a review. Crit. Rev. Environ. Sci.
Technol. 45, 429–471. https://doi.org/10.1080/10643389.2013.866622.