Colloidal rare earth elements in a boreal river: Changing sources and distributions during the spring flood

doi:10.1016/j.gca.2006.04.021

Geochimica et Cosmochimica Acta

Volume 70, Issue 13, 1 July 2006, Pages 3261-3274

https://doi.org/10.1016/j.gca.2006.04.021 Get rights and content

Abstract

Variations in the physico-chemical speciation of the rare earth elements (REE) have been investigated in a subarctic boreal river during an intense spring flood event using prefiltered (<100 μm) samples, cross-flow (ultra)filtration (CFF), flow field-flow fractionation (FlFFF), and diffusive gradients in thin films (DGT). This combination of techniques has provided new information regarding the release and transport of the REE in river water. The colloidal material can be described in terms of two fractions dominated by carbon and iron, respectively. These two fractions, termed colloidal carrier phases, showed significant temporal changes in concentration and size distribution. Before the spring flood, colloidal carbon concentrations were low, the colloids being dominated by relatively large iron colloids. Colloidal concentrations increased sharply during the spring flood, with smaller carbon colloids dominating. Following the spring flood, colloidal concentrations decreased again, smaller carbon colloids still dominating. The REE are transported mainly in the particulate and colloidal phases. Before the spring flood, the REE composition of all measured fractions was similar to local till. During the spring flood, the REE concentrations in the colloidal and particulate fractions increased. The increase was most marked for the lighter REE, which therefore showed a strong enrichment when normalized to local till. Following the spring flood, the REE concentrations decreased again and reverted to a distribution similar to local till. These changes in the concentration and distributions of carbon iron and REE are interpreted in terms of changing hydrological flow paths in soil and bedrock which occur during the spring flood.

Introduction

The rare earth elements (REE) can be used as tracers to study processes that control the composition of natural waters. A topic of fundamental importance to geochemistry is elemental fractionation between continents and oceans. Since the dominant source for rare earth elements in seawater is continental, terrestrial low temperature reactions and processes such as weathering, surface chemistry, river transport and estuarine cycling have a great impact on the fate of the REE.

It has been demonstrated that the REE are mobile during weathering in cold climates (Floss and Crozaz, 1991) and temperate regions (Öhlander et al., 1991), and that an elemental fractionation occurs (Öhlander et al., 1996), causing the light REE (LREE) to be released to a greater extent than the heavy REE (HREE). This elemental fractionation is suggested to be a result of selective weathering of minerals in the upper part of soil profiles (E-horizon), and can be observed in the river REE distributions (Ingri et al., 2000, Andersson et al., 2001). This may cause REE entering the river to be significantly fractionated relative to the local bedrock sources (Öhlander et al., 1996, Viers et al., 1997, Land et al., 1999a, Ingri et al., 2000, Andersson et al., 2001).

In addition to selective weathering, elemental fractionation may also occur during aqueous transport, where natural particles and colloids are of great importance. The extensive chemical changes occurring at the fresh water–salt water interface have focused much attention on the role of estuaries in modifying the land to ocean transport of REE (e.g., Elderfield et al., 1990, Sholkovitz, 1993, Sholkovitz et al., 1999, Sholkovitz and Szymczak, 2000). Several studies have also addressed the physico-chemical speciation of REE in continental waters (e.g., Goldstein and Jacobsen, 1987, Goldstein and Jacobsen, 1988, Elderfield et al., 1990, Sholkovitz, 1992, Sholkovitz, 1995, Ingri et al., 2000, Tang and Johannesson, 2003). In particular, the strong association of REE with (colloidal) organic matter in terrestrial waters has been demonstrated in several ultrafiltration investigations (Tanizaki et al., 1992, Viers et al., 1997, Dupré et al., 1999, Ingri et al., 2000).

The Kalix River, an unregulated pristine boreal river system, has previously been well characterized in several studies on major and trace elements and isotopes (Pontér et al., 1990, Öhlander et al., 1991, Öhlander et al., 1996, Öhlander et al., 2000, Ingri and Widerlund, 1994, Ingri et al., 1997, Ingri et al., 2000, Ingri et al., 2005, Porcelli et al., 1997, Land and Öhlander, 1997, Land and Öhlander, 2000; Andersson et al., 1998, Andersson et al., 2001, Land et al., 1999a, Land et al., 1999b, Land et al., 2000a, Land et al., 2000b, Dahlqvist et al., 2004). The study by Ingri et al. (2000) showed that transport of REE in the Kalix River was closely associated with colloidal particles rich in C, Fe and Al. Significant increases in DOC concentrations (a factor of 10) coincided with high discharge during snow melt in spring. This behavior has been observed in several rivers at high latitudes (e.g., Craig and McCart, 1975, Cauwet and Sidorov, 1996, Rember and Trefry, 2004) and is believed to be an effect of percolating melt water or rising levels of shallow ground water in top organic-rich soil horizons during spring.

Normalized REE patterns in the Kalix River reveal that the colloidal fraction is enriched in LREE during spring flood compared to periods with low discharge. The study by Ingri et al. (2000) also suggested that two different major carrier phases for the REE are present in the river; Fe-oxyhydroxides, and organic material with associated Al and Fe. The colloidal phase is particularly difficult to assess due to its size range (nominally 1 nm–1 μm) and chemical and physical heterogeneity (Gustafsson and Gschwend, 1997). Since none of the available methods for determination of colloids are without potential artifacts, researchers studying this fraction are faced with the problem of separating effects due to natural changes from artifacts introduced while applying the method. In order to obtain a more complete picture of the sub-micrometer size fractionation, the study therefore combines cross-flow filtration (CFF) or ultrafiltration (UF), flow field-flow fractionation (FlFFF), and diffusive gradients in thin films (DGT).

The two major aims of this study on the Kalix River are (1) to verify and characterize the two proposed colloidal carrier phases (Ingri et al., 2000), and (2) to establish normalized REE patterns in the colloidal and DGT-labile fractions. This study on REE fractionation is part of the ‘Kalix 2002’ project, in which we have studied temporal changes in the geochemistry of colloids in the Kalix River from winter conditions, through the spring flood event and into summer conditions. Other papers from this project include a study of calcium association with colloids (Dahlqvist et al., 2004), together with studies of multiway analysis of organic matter fluorescence, and of the temporal changes of trace metals associated with colloidal material, which will be presented elsewhere.

Section snippets

Sampling area

The Kalix River in northern Sweden (Fig. 1) is part of a pristine river system which also includes the Torne River (Dynesius and Nilsson, 1994); no man-made dams or other constructions influence the water discharge. The mean annual discharge is 296 m³ s⁻¹, with low flows below 100 m³ s⁻¹ during winter, and high flows up to 1600 m³ s⁻¹ during the spring flood. The drainage area is ∼2.4 × 10⁴ km² and stretches from the Scandinavian part of the Caledonian mountain range in the northwest to the Gulf of

Results and discussion

The hydrochemistry of Kalix River during the study period was similar to that observed previously in the river, except for the timing of the spring flood which was unusually early (by about 2 weeks), with the peak flood occurring during the period 1–6 May. Over a 10-day period, the discharge increased from <100 to 1640 m³ s⁻¹, accompanied by sharp changes in pH and concentrations of organic carbon and iron (Fig. 3). Fig. 3 also shows the strong increase in hours of sunlight and water temperature

Conclusions

Results from prefiltered samples, ultrafiltration fractions, DGT deployments, and FlFFF experiments suggest that the REE mainly are transported in the colloidal (1 kDa–0.22 μm) and particulate (>0.22 μm) phases in the Kalix River. The different methods show that REE concentrations increase during spring flood, and that the increase is more pronounced for LREE than for HREE.

Two different colloidal carrier phases, different in size and chemical composition were identified with FlFFF. A small (∼3 nm)

Acknowledgments

The authors thank Anders Henriksson, Johan Gelting-Nyström, Helena Skoglund and Jerry Forsberg at Luleå University of Tech. for their assistance during collection and processing of samples. We are grateful to Rickard Hernell and Ilia Rodushkin at Analytica Corp. for help during sample preparation, analysis and setup of the MS used for FlFFF. We thank Hao Zhang for fruitful discussions on the interpretation of DGT results. This work is supported by Swedish Research Council Grants G

References (60)

P.S. Andersson et al.
Particle transport of ²³⁴U–²³⁸U in the Kalix River and the Baltic Sea
Geochim. Cosmochim. Acta
(1998)
P.S. Andersson et al.
The isotopic composition of Nd in a boreal river: a reflection of selective weathering and colloidal transport
Geochim. Cosmochim. Acta
(2001)
G. Cauwet et al.
The biogeochemistry of the Lena Rvier: organic carbon and nutrients distribution
Mar. Chem.
(1996)
R. Dahlqvist et al.
Performance of the diffusive gradients in thin films technique for measuring Ca and Mg in freshwater
Anal. Chim. Acta
(2002)
R. Dahlqvist et al.
Association of calcium with colloidal particles and speciation of calcium in the Kalix and Amazon rivers
Geochim. Cosmochim. Acta
(2004)
H. Elderfield et al.
The rare-earth elements in rivers, estuaries, and coastal seas and their significance to the composition of ocean waters
Geochim. Cosmochim. Acta
(1990)
C. Floss et al.
Ce anomalies in the LEW85300 meucrite: evidence for REE mobilization during Antarctic weathering
Earth Planet. Sci. Lett.
(1991)
G. Gaal et al.
An outline of the Precambrian evolution of the Baltic Shield
Precambrian Res.
(1987)
S.J. Goldstein et al.
The Nd and Sr isotopic systematics of river water dissolved material: implications for the sources of Nd and Sr in seawater
Chem. Geol. (Isot. Geosci. Sect.)
(1987)
S.J. Goldstein et al.
Rare earth elements in river waters
Earth Planet. Sci. Lett.
(1988)

E.R. Sholkovitz

Chemical evolution of rare-earth elements—fractionation between colloidal and solution phases of filtered river water

Earth Planet. Sci. Lett.

(1992)

E.R. Sholkovitz

The geochemistry of rare-earth elements in the Amazon River estuary

Geochim. Cosmochim. Acta

(1993)

Cited by (115)

Organic-mineral colloids regulate the migration and fractionation of rare earth elements in groundwater systems impacted by ion-adsorption deposits mining in South China
2024, Water Research
Ion-adsorption rare earth element (REE) deposits distributed in the subtropics provide a rich global source of REEs, but in situ injection of REEs extractant into the mine can result in leachate being leaked into the surrounding groundwater systems. Due to the lack of understanding of REE speciation distribution, particularly colloidal characteristics in a mining area, the risks of REEs migration caused by in situ leaching of ion-adsorption REE deposits has not been concerned. Here, ultrafiltration and asymmetric flow field-flow fractionation coupled with inductively coupled plasma mass spectrometry (AF4-ICP-MS) were integrated to characterize the size and composition of REEs in leachate and groundwater from mining catchments in South China. Results show that REEs were associated with four fractions: 1) the <1 kDa fraction including dissolved REEs; 2) the 1 – 100 kDa nano-colloidal fraction containing organic compounds; 3) the 100 kDa – 220 nm fine colloids including organic-mineral (Fe, Mn and Al (oxy)hydroxides and clay minerals); 4) the >220 nm coarse colloids and acid soluble particles (ASPs) comprising minerals. Influenced by the ion exchange effect of in situ leaching, REEs in leachate were mostly dissolved (79 %). The pH of the groundwater far from the mine site was increased (5.8 – 7.3), the fine organic-mineral colloids (46 % – 80 %) were the main vectors of transport for REEs. Further analysis by AF4 revealed that the fine colloids can be divided into mineral-rich (F1, 100 kDa – 120 nm) and organic matter-rich (F2, 120 – 220 nm) populations. The main colloids associated with REEs shifted from F1 (64 % ∼ 76 %) to F2 (50 % ∼ 52 %) away from the mining area. For F1 and F2, the metal/C molar ratio decreased away from the mining area and middle to heavy REE enrichment was presented. According to the REE fractionation, organic matter was the predominant component capable of binding REEs in fine colloids. Overall, our results indicate that REEs in the groundwater system shifted from the dissolved to the colloidal phase in a catchment affected by in situ leaching, and organic-mineral colloids play an important role in facilitating the migration of REEs.
Diagenetic imprints in the reservoirs of Agbada Formation, Niger Delta Basin: Implication for reservoir quality and appraisal of calcite cement
2024, Marine and Petroleum Geology
The main exploration targets for hydrocarbons in the Niger Delta Basin are within the Agbada sandstone reservoirs. The current study utilizes a multitechnical approach (SEM-MLA, SEM-EDS, XRD and LA-ICP-MS) to evaluate the roles of framework composition and diagenetic processes, particularly calcite cementation on reservoir quality in Miocene intervals of the Agbada Formation sandstone, Niger Delta Basin. Petrographic analysis reveals that the sandstones are feldspatho-quartzose to quartzose sandstones, with a preponderance of quartzose sandstones. The overall diagenetic imprints suggest minimal compaction and lack of pervasive cement, which have contributed significantly to the high porosity. Although the degree of mechanical compaction is generally low, it has comparatively exercised a better control on porosity than the growth of cement, excluding some intervals with considerable cementation (siderite and ferroan calcite) where compactional porosity-loss was minimal. The occurrence of ferroan calcite cement (CaCO₃ = 95.9 ± 0.4, MgCO₃ = 0.8 ± 0.1, FeCO₃ = 2.9 ± 0.2 mol%) is generally small, probably discontinuous and unlikely to represent significant barriers to fluid flow with their localized influence, but understanding their locally significant influence may still contribute to building good reservoir models. Petrographic observations suggest that the precipitation of “poikilitic” calcite was early, but the shale normalized REY patterns do not mimic seawater patterns (Y/Ho ratios; 36.3 ± 1.6) due to possible contributions from clastic or detrital grains and/or incorporation of particulate matter that can preferentially scavenge LREEs from the overlying water column. A possible concern during reservoir production is the occurrence of clay mineral matrix (mostly kaolinitic) anticipated to cause the production of fines and formation damage, but this can be mitigated with proper reservoir management.
Extraction and characterization of pore water from contaminated soils
2024, Environmental Geochemistry: Site Characterization, Data Analysis, Case Histories, and Associated Health Issues
Chemical elements that are either present naturally in the soil or introduced by pollution are more usefully estimated in terms of “availability” of the element since it is this property that can be related to mobility and uptake by plants. A good estimate of the immediately available fraction can be achieved by measuring the concentration, or activity, of chemical species in soil pore water. Current analytical techniques enable the application of this approach to trace elements, such as plant and animal micronutrients, and those defined as potentially toxic elements (PTE) in environmental studies. A complete chemical analysis of soil pore water represents a powerful diagnostic tool for the interpretation of many soil chemical phenomena relating to soil fertility, mineralogy, and environmental fate. This chapter describes some of the current methodologies used to extract soil pore water. In particular, five laboratory-based methods, (i) high-speed centrifugation-filtration, (ii) low (negative-) pressure Rhizon samplers and passive diffusion samplers (Micro-Dialysis Probes MDPs), (iii) high-pressure soil squeezing, (iv) equilibration of dilute soil suspensions, and (v) Diffusive Gradients in Thin-films (DGT), are described and discussed in detail. Several operational factors are presented: pressure applicable (i.e., pore size accessed), moisture prerequisites of the soil, pore water yield, efficiency, duration of extraction, materials, and possible sources of contamination for micronutrient and PTE studies. There is also consideration of the advantages and disadvantages of the methods, including costs and material availability.
Export of organic carbon, nutrients and metals by the mid-sized Pechora River to the Arctic Ocean
2023, Chemical Geology
In contrast to good knowledge of export fluxes of carbon and metals from mainland to the Arctic Ocean by large Arctic Rivers, information on mid-sized rivers is limited, which prevents determining current status and foreseeing future changes in riverine export induced by climate change. Here we focused on one of the ‘middle eight’ Arctic rivers. The Pechora River (S_watershed = 322,000 km²) is the second largest European Arctic river draining through boreal forest and peatlands with partial (∼ 40%) permafrost coverage. Over 4 consecutive years (2015–2019), we measured weekly to monthly concentrations of carbon, major nutrients, and 40 major and trace elements in filtered (< 0.45 μm) river water at the terminal gauging stations for the Pechora River. The dependences between dissolved element concentration and river discharge over the full period of observation revealed 3 groups of major and trace solutes according to their seasonal behavior. Group 1 was comprised of Dissolved Inorganic Carbon (DIC), major anions (Cl, SO₄), alkalis (Li, Na, K, Rb), alkaline-earth metals (Mg, Ca, Sr, Ba), elements present in the form of labile anions and neutral molecules (B, Si, Ge, Mo, Sb) and U. It demonstrated minimal concentrations during spring flooding and autumn high flow events, and maximal concentrations during winter base flow. Concentrations of these element negatively correlated with discharge. These elements primarily reflected the dominant sedimentary lithology of the Pechora catchment and were controlled by influx of underground waters hosted in carbonate rocks. Group 2 included DOC and low-mobility trivalent and tetravalent hydrolysates (Be, Al, Ga, Y, REEs, Ti, Zr, Hf, Th) and some trace metals (V, Cr, Cs, Nb). The transport of these elements to the river from the catchment likely occurred via surface and shallow subsurface waters due to leaching from organic-rich litter layer, and their concentrations positively correlated with discharge. Finally, group 3 involved major- (P, N, K) and micro-nutrients (Fe, Mn, Cu, Zn, Mo) and demonstrated features of both the 1st and 2nd groups as these elements originated from surface (organic-rich) and underground (sedimentary rock-hosted) sources and were removed by autochthonous biotic processes in the water column. As a result, the concentration of these nutrients exhibited a winter maximum and a summer minimum.
Similar to other Arctic and subarctic rivers, the spring flood, which lasted from May to June, accounted for >50% of the annual export of DOC and generally insoluble trace metals. However, the winter baseflow provided sizable proportions of mobile elements such as DIC, major anions, cations, alkaline-earth elements, trace oxyanions and uranium. The size distribution of riverine solutes demonstrated 20–50% of DOC in the colloidal (3 kDa – 0.45 μm) form, dominance of low molecular forms (< 3 kDa) of soluble elements (alkali, alkaline earths, anions and neutral molecules) and essentially colloidal status of divalent metals (Cu, Zn, Cd, Pb, Ni), trivalent and tetravalent hydrolysates and U (from 30 to 93%). Seasonal variations in colloidal fractions of riverine solutes were consistent with the main mechanisms of element mobilization from the watershed to the Pechora River, as reflected in the 3 groups of elements according to their concentration – discharge relationship. Based on the present study and results of other mid-sized Arctic rivers (Severnaya Dvina, Taz), we demonstrate that mean discharges and solute concentrations in August can be used to adequately (± 20%) approximate mean annual values, hence allowing annual flux assessment via single month sampling campaign. A comparison of contemporary yields of Ca, Mg, Cl, SO₄, DIC and DOC against historical data for the Pechora River basin demonstrated reasonable agreement within the inter-annual variations that is consistent with long-term data series on other Arctic rivers.
Rare earth elements (REEs) behavior in a large river across a geological and anthropogenic gradient
2023, Geochimica et Cosmochimica Acta
To meet the strong demand for green and digital technologies, the production of rare earth elements (REEs) has increased significantly, yet the fate of REEs (Sc, Y, La, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) in large rivers with multiple hydrological, geochemical, and anthropogenic gradients in their catchments remains poorly documented. Here we evaluated the spatial and temporal evolution of the concentrations, patterns, and speciation of REEs in surface waters along a 550 km transect of the St. Lawrence River (SLR) (Canada) and its tributaries from 2017 to 2020. Particulate and filtered (<0.45 µm) REEs concentrations increased downstream, suggesting that REEs found in the SLR predominantly originate from nearby tributaries, especially those draining the REE-rich Canadian Shield geological formation. Moreover, filtered concentrations were significantly higher in years that experienced intense spring freshets, indicating greater influence of runoff and surface soil processes on water chemistry of the SLR. Correlations with terrestrial-DOM, WHAM speciation modelling, and standardized patterns showed REEs speciation dominated by organic colloids, even in alkaline sites with low DOC and low terrestrial inputs. Strong positive correlations with Fe and Al suggested co-transport on organic matter and adsorption on oxyhydroxides. Compared to the influence of water chemistry and geology, the overall influence of anthropogenic inputs was generally low, although we observed an influence of increased erosion on particulate REEs and we measured anthropogenic gadolinium (Gd) at the Montréal wastewater outlet. Our results provide an integrative understanding of the fate of REEs through a geochemical and landscape perspective in one of the largest rivers in North America.
Loss of subsurface particulate and truly dissolved phosphorus during various flow conditions along a tile drain–ditch–brook continuum
2023, Science of the Total Environment
Subsurface losses of colloidal and truly dissolved phosphorus (P) from arable land can cause ecological damage to surface water. To gain deeper knowledge about subsurface particulate P transport from inland sources to brooks, we studied an artificially drained lowland catchment (1550 ha) in north-eastern Germany. We took daily samples during the winter discharge period 2019/2020 at different locations, i.e., a drain outlet, ditch, and brook, and analyzed them for total P (TP_unfiltered), particulate P >750 nm (TP_{>750 nm}), colloidal P (TP_colloids), and truly dissolved P (truly DP) during baseflow conditions and high flow events. The majority of TP_unfiltered in the tile drain, ditch, and brook was formed by TP_{>750 nm} (54 to 59 %), followed by truly DP (34 to 38 %) and a small contribution of TP_colloids (5 to 6 %). During flow events, 63 to 66 % of TP_unfiltered was present as particulate P (TP_{>750 nm} + TP_colloids), whereas during baseflow the figure was 97 to 99 %; thus, truly DP was almost negligible (1 to 3 % of TP_unfiltered) during baseflow. We also found that colloids transported in the water samples have their origin in the water-extractable nanocolloids (0.66 to 20 nm) within the C horizon, which are mainly composed of clay minerals. Along the flow path there is an agglomeration of P-bearing nanocolloids from the soil, with an increasing importance of iron(III) (hydr)oxides over clay particles. Event flow facilitated the transport of greater amounts of larger particles (>750 nm) through the soil matrix. However, the discharge did not exhaust colloid mobilization and colloidal P was exported through the tile-drainage system during the complete runoff period, even under baseflow conditions. Therefore, it is essential that the impact of rainfall intensity and pattern on particulate P discharge be considered more closely so that drainage management can be adjusted to achieve a reduced P export from agricultural land.

View all citing articles on Scopus

¹: Present address: University of Oxford, Department of Earth Sciences, Parks Road, Oxford OX1 3PR, UK.

View full text

Colloidal rare earth elements in a boreal river: Changing sources and distributions during the spring flood

Abstract

Introduction

Section snippets

Sampling area

Results and discussion

Conclusions

Acknowledgments

Geochim. Cosmochim. Acta

Geochim. Cosmochim. Acta

Mar. Chem.

Anal. Chim. Acta

Geochim. Cosmochim. Acta

Geochim. Cosmochim. Acta

Earth Planet. Sci. Lett.

Precambrian Res.

Chem. Geol. (Isot. Geosci. Sect.)

Earth Planet. Sci. Lett.

Water Res.

Chem. Geol.

Geochim. Cosmochim. Acta

Appl. Geochem.

Chem. Geol.

Chem. Geol.

Appl. Geochem.

Chem. Geol.

Appl. Geochem.

Anal. Chim. Acta

Geochim. Cosmochim. Acta

J. Solid State Chem.

Appl. Geochem.

Geochim. Cosmochim. Acta

Chem. Geol.

Geochim. Cosmochim. Acta

Geochim. Cosmochim. Acta

Anal. Chim. Acta

Earth Planet. Sci. Lett.

Geochim. Cosmochim. Acta