Gadolinium

    no group (f-block) · period 6 · f-block · lanthanide

    minorGadolinium is the lanthanide with a real water story: the chelated contrast agents of magnetic resonance imaging pass patients, sewers, activated sludge, rivers, bank filtration and waterworks almost unchanged, so the positive gadolinium anomaly has become the tracer of treated sewage in rivers and of sewage in tap water. It is still unregulated, so the verdict stays minor.

    Typical wastewaters

    • hospital effluent and municipal sewage (MRI contrast agents) intact Gd(III) polyaminocarboxylate chelates, anionic (linear, gadopentetate type) or neutral (macrocyclic), truly dissolved below 10 kDa; influent and effluent of metropolitan sewage plants enriched, sludge not four full scale metropolitan plants; a large plant's biosolids were not enriched although its effluent was
    • water recycling (reverse osmosis concentrate) the rejected chelate concentrated 6 to 7 fold in the brine, 2.6 nmol/kg against 0.39 nmol/kg in the effluent feed proposed as a tracer of the concentrate's fate
    • coal mine drainage (acid) Gd³⁺ and the GdSO₄⁺ ion pair dissolved at low pH in sulfate rich drainage; total rare earths averaged 282 µg/L over 141 Appalachian sites, per-element values not read co-precipitates into the iron, aluminium and manganese hydroxide sludge when the drainage is neutralised

    1 · Identity

    Symbol, number
    Gd, 64
    Oxidation states in water
    +3 only (Gd³⁺); in water almost all anthropogenic gadolinium is Gd(III) held in linear or macrocyclic polyaminocarboxylate chelates
    Note
    Free Gd³⁺ is toxic and is what the chelate exists to prevent; the chelates are stable enough to survive the water cycle, which is the whole problem.

    2 · Occurrence in water

    Natural sources
    Weathering of monazite and bastnaesite; geogenic Gd in the lower Rhine is 1.5 to 3.6 ng/kg, in seawater 1.53 to 9.65 pmol/L. Middle lanthanide.
    Anthropogenic sources
    Gadolinium-based contrast agents excreted in urine within hours of an MRI examination, reaching sewers, hospital effluent, sewage treatment plants and receiving water. Bau and Dulski (1996) showed positive Gd anomalies in rivers of densely populated central Europe and North America and none in thinly populated Sweden and Japan, almost entirely in the dissolved fraction below 0.2 µm, and traced them to gadopentetate. Since then: anthropogenic Gd of 7.67 to 25.0 ng/kg on a geogenic 1.5 to 3.6 ng/kg in the lower Rhine (up to 99 percent anthropogenic in some samples, up to 730 kg a year to the North Sea); an order of magnitude rise in San Francisco Bay from 8.27 to 112 pmol/kg over two decades; 85 to 99 percent anthropogenic Gd in Berlin and Dusseldorf tap water produced by bank filtration and 91 percent in Munich tap water from Alpine valley groundwater; a 1.5 to 11.5 fold rise in Berlin tap water between 2009 and 2012. Sewage treatment plant effluents show the anomaly wherever the catchment has medical facilities; the one plant serving 1,200 people with none showed no anomaly.
    matrixtypical rangenote
    surface water, lower Rhine (river-km 811)7.67 to 25.0 ng/kg
    one river; anthropogenic share varies with flow and season
    anthropogenic Gd in three campaigns 2008 to 2010, on geogenic Gd of 3.61, 1.52 and 1.60 ng/kg
    estuary and coastal water, San Francisco Bay8.27 to 112 pmol/kganthropogenic Gd, early 1990s to mid 2010s; highest in the southern reach near hospitals; reaches the Pacific coast
    seawater, open ocean1.53 to 9.65 pmol/Lone stationwestern Pacific, 3 to 5663 m, geogenic
    sewage effluent used for water recycling0.39 nmol/kgone plantanthropogenic Gd in the feed of an advanced water treatment plant; RO permeate 0.59 pmol/kg; RO concentrate 2.6 nmol/kg
    drinking water85 to 99 percent anthropogenic percent of total Gd
    concentrations are in the full text, not the abstract read
    Berlin and Dusseldorf tap water from bank filtration; 91 percent in Munich; present in all six German cities sampled and unchanged in the soft drinks made from the tap water

    3 · Speciation

    Trivalent Gd³⁺ throughout the natural pH and Eh range. Below about pH 6, and in sulfate rich water such as mine drainage, the free ion and the GdSO₄⁺ ion pair dominate (sulfate complexes above 90 percent at pH 3.1 to 3.4 in the EPA groundwater study). From neutral to alkaline pH the carbonate complexes GdCO₃⁺ and Gd(CO₃)₂⁻ take over, and they bind the heavy lanthanides more strongly than the light ones (at pH 5.1 to 6.1 the carbonate share rose from 5.0 percent for La to 19.2 percent for Lu). Phosphate, carbonate and fluoride solids and sorption to iron and manganese oxides cap the dissolved concentration at ng/L in oxic neutral water, and the light lanthanides ride partly on colloids. Anthropogenic gadolinium is not Gd³⁺ at all but the intact chelate: an anionic (linear, gadopentetate type) or neutral (macrocyclic) complex that is not particle reactive, stays in the truly dissolved pool below 10 kDa, is not taken into mussel shells, and behaves conservatively through estuaries and aquifers. Speciation modelling of effluent affected river water shows that even geogenic REE stay dissolved there because phosphate and carbonate complexes dominate over the free ion.

    conditiondominant speciesnote
    acid mine drainage and acidic groundwater, pH below 5Gd³⁺, GdSO₄⁺the most mobile state; dissolved total REE in coal mine drainage averaged 282 µg/L (Part 2 survey), against ng/L in neutral water
    neutral to alkaline groundwater and river water, pH 7 to 9GdCO₃⁺, Gd(CO₃)₂⁻; a small free ion fraction; part of the load on colloids below 0.2 µmthe carbonate complexes keep the element in solution and make the heavy lanthanides relatively more mobile
    phosphate rich water, treated lake sedimentsGdPO₄ (s) as a hydrated phosphateREE phosphate solubility products can be as low as 10⁻25 (RIVM report citing Liu and Byrne 1997)
    sewage effluent, rivers, groundwater and tap waterintact Gd chelates (linear polyaminocarboxylates and macrocyclic ligands); transmetallation and dechelation slowstability depends on the ligand: linear agents result in more gadolinium retention in the body than macrocyclic ones (FDA); under UV one linear agent degraded to 3 percent under UV in 300 min in purified water, slower in drinking and surface water, with Gd containing transformation products, while three other agents were stable to UV
    Solubility
    Controlled by the phosphate, carbonate and fluoride solids and by sorption; the RIVM report puts total dissolved lanthanum and cerium in Dutch surface water at about 1 µg/L or less and the free ion fraction at 0.3 fM to 9 pM. No solubility product for a gadolinium solid was read this session.
    Hydrolysis
    Hydrolysis of Gd³⁺ is minor in natural water; the hydroxide Gd(OH)₃ forms only at high pH, so lime or caustic neutralisation of acid mine water strips the element with the iron and aluminium hydroxides rather than as its own hydroxide (no hydrolysis constant read this session).
    Complexation
    The whole story: polyaminocarboxylate chelates with conditional stability high enough to survive sewage treatment, bank filtration and years of aquifer transit; carbonate and phosphate complexes for the geogenic fraction. Brunjes and Hofmann (2020) warn that the chelates do degrade, ligand by ligand and with UV end-of-pipe treatment, and that degradation products in raw water with a high recycled wastewater share raise the health question.
    Precipitates
    For geogenic Gd: GdPO₄ (hydrated), Gd₂(CO₃)₃, Gd(OH)₃ under lime. The chelates form no precipitate in treatment.
    GdX3++COX3X2GdCOX3X+\ce{Gd^3+ + CO3^2- -> GdCO3^+}
    circumneutral pH; the first carbonate complex
    GdX3++2COX3X2Gd(COX3)X2X\ce{Gd^3+ + 2 CO3^2- -> Gd(CO3)2^-}
    alkaline pH; the dicarbonate anion, strongest for the heavy lanthanides
    GdX3++SOX4X2GdSOX4X+\ce{Gd^3+ + SO4^2- -> GdSO4^+}
    acid, sulfate rich water; above 90 percent of the dissolved element at pH 3.1 to 3.4
    GdX3++POX4X3GdPOX4(s)\ce{Gd^3+ + PO4^3- -> GdPO4 (s)}
    phosphate present; the solid that caps solubility and that phosphate binders rely on

    4 · Role in treatment

    as a problem
    passage through sewage treatment
    the chelate stays in the liquid phase through primary, secondary and biosolids processing; influent and effluent are enriched in Gd, sludges are not
    four full scale metropolitan plants (Verplanck 2010); a large plant's biosolids not enriched although its effluent was (Verplanck 2005)
    passage through bank filtration and waterworks into tap water
    conservative transport from river to production wells; wells take years to decades, so tap water Gd lags river Gd and will keep rising
    Berlin western districts strongly affected, eastern districts not, following the water sources; Munich Alpine groundwater 91 percent anthropogenic
    UV disinfection
    some linear agents photodegrade with release of Gd(III) ions and formation of more polar Gd containing transformation products by O- and N-dealkylation
    a concern for end-of-pipe UV and for potable reuse (Brunjes and Hofmann)
    RO concentrate
    the chelate is rejected almost completely, so it concentrates 6 to 7 fold in the brine
    2.6 nmol/kg in the concentrate against 0.39 nmol/kg feed; proposed as a tracer of the concentrate's fate
    as a reagent
    tracer of treated sewage and of river water in aquifers
    the anthropogenic Gd anomaly is conservative, cheap to measure by ICP-MS beside organic wastewater compounds, and its temporal variation in river water reads river to aquifer travel times
    requires the shale-normalised interpolation of geogenic Gd from its neighbours; at low anomalies the choice of calculation method changes the result

    5 · Removal and control

    reverse osmosis
    size and charge rejection of the chelate
    advanced water treatment plant producing purified recycled water from effluent; coagulation and microfiltration ahead of it removed nothing
    Efficiency
    99.85 percent (0.39 nmol/kg to 0.59 pmol/kg anthropogenic Gd)
    Interferences
    the rejected Gd ends in the concentrate at 2.6 nmol/kg
    activated sludge and biosolids processing
    none for the chelate; it stays in the liquid
    four metropolitan plants
    Efficiency
    negligible
    coagulation and microfiltration
    the chelate is neither particle reactive nor colloidal
    same plant as the RO
    Efficiency
    negligible
    bank filtration and aquifer passage
    conservative transport
    Berlin, Dusseldorf, Munich supplies
    Efficiency
    negligible
    source control: urine collection after MRI
    keep the excreted agent out of the sewer and recover the gadolinium
    proposed, not practised at scale
    Efficiency
    not quoted

    6 · Analytics

    methodstandarddetection limitnote
    ICP-MSISO 17294-2:2023not read; the standard covers drinking, surface, ground and waste watergadolinium is in the element list of ISO 17294-2 (catalogue abstract)
    ICP-MS after preconcentrationresearch methods: chelating resin columns or magnesium hydroxide co-precipitation, then quadrupole or high resolution ICP-MS0.012 to 0.98 pmol/L (procedural blanks 0.041 to 3.60 pmol/L) for the 14 lanthanides pooled, magnesium hydroxide co-precipitation with a factor of about 200; terbium has been measured at 7 fmol/kg in recycled waterthe only way to reach the ng/L and sub-ng/L levels of rivers and seawater; the shale-normalised pattern, not the single concentration, is what the geochemists read
    HILIC-ICP-MS and HILIC-ESI-MSresearch method (Birka et al. 2016)not readseparates the individual chelates and their transformation products
    Sampling pitfalls
    Filtration defines the result: 0.2 or 0.45 µm filtrates still carry colloid bound lanthanides, and only ultrafiltration (10 kDa) separates the truly dissolved pool, which matters most for the light lanthanides. Acidify after filtration. Report the shale-normalised pattern so that anthropogenic anomalies (gadolinium, lanthanum, samarium) are visible. Anthropogenic Gd is a difference: total Gd minus geogenic Gd interpolated from the shale-normalised neighbours (Sm, Eu, Tb, Dy); the interpolation method matters when the anomaly is small (Brunjes and Hofmann). Speciation of individual contrast agents needs hydrophilic interaction chromatography coupled to ICP-MS or electrospray MS.

    7 · Regulatory limits

    Limits change, and many are set locally. Treat these as the published values to start from, not as your compliance target: check the standard in force at your site and the numbers written into your own permit.

    drinking water
    bodylimitnote
    US EPA National Primary Drinking Water Regulationsnot regulated no lanthanide or rare earth element in the NPDWR table
    discharge
    bodylimitnote
    EU CWW BAT-AEL (Decision 2016/902), BAT 12not set Tables 1 to 3 carry TOC, COD, TSS, TN, Ninorg, TP, AOX, Cr, Cu, Ni, Zn; no rare earth element

    8 · Health and environmental effects

    Toxicity
    No water guideline. In patients: gadolinium from contrast agents is retained in the body, including the brain, for months to years (FDA class warning of December 2017); the one known adverse effect is nephrogenic systemic fibrosis in a small subgroup of patients with pre-existing kidney failure; linear agents retain more gadolinium than macrocyclic ones. Free Gd³⁺ is the toxic form; the drinking water question is degradation products in raw water with a high recycled wastewater share (Brunjes and Hofmann), not the ng/L of intact chelate.
    Bioaccumulation
    Field bioconcentration factors in Rhine estuary amphipods fall from the light to the heavy lanthanides (La 28,840 to Lu 4,786; Gd 13,183); carp muscle takes up little (BCF 0.22 to 1.10 for Ce, La, Nd, Pr, Sm) while internal organs reach 634 to 978. Anthropogenic La and Sm in the Rhine are taken into mussel shells, anthropogenic Gd is not, so speciation decides bioavailability.
    Ecotoxicity
    RIVM (2000) environmental risk limits, not discharge limits: MPC 7.1 µg/L in fresh surface water and 0.85 µg/L in salt water (negligible concentrations 0.40 and 0.34 µg/L), derived as lowest LC₅₀ divided by 1000 plus a background set at the detection limit (0.33 µg/L). Acute data behind it: Daphnia magna 48 h EC₅₀ 6.8 mg/L, zebrafish 96 h LC₅₀ 19 mg/L; chronic: no chronic test listed. Field bioconcentration factor in amphipods (porewater basis) 13,183. Anthropogenic Gd is not taken into freshwater mussel shells, unlike anthropogenic La and Sm, so the RIVM figures, which are for dissolved gadolinium salts, do not describe the chelate.

    Flags

    • Schmidt 2019 tap water concentrations are in the full text; only the percentages of the abstract are written.
    • The Tepe 2014 figure (1.5 to 11.5 fold rise) is taken from the abstract as summarised in search results, not from the paper.
    • Bau and Dulski 1996 is cited from its abstract.
    • The FDA communication names contrast agent brands; only the linear and macrocyclic classes are written.
    • The Hatje paper names a chelating resin brand; it is written as a chelating resin.
    • The RIVM MPC (7.1 µg/L fresh, 0.85 µg/L salt) is for gadolinium salts, LC₅₀ divided by 1000, and does not apply to the chelates.

    Gaps

    • WHO GDWQ and EU DWD 2020/2184 Annex I were not read this session (eur-lex returned the articles without the annexes), so their absence of a rare earth parameter is not asserted here; only the US EPA table was read.
    • No GCC discharge standard was read; no GCC row is written.
    • World average river water concentrations (Gaillardet et al. 2003) were not reachable; river figures come from the Rhine and Dutch waters only.
    • No source read this session gives municipal wastewater concentrations for this element beyond the gadolinium literature.
    • ICP-MS oxide interference corrections between lanthanides are standard practice but no method text describing them was read, so none is written.
    • No stability constants or solubility products for this element's carbonate, phosphate or hydroxide were read; the speciation is qualitative, taken from the EPA groundwater study and the RIVM report.
    • Hospital effluent and sewage influent Gd concentrations were not read as numbers.
    • Removal by nanofiltration, ozonation and activated carbon was not read as numbers; the Brunjes and Hofmann review covers them but only its abstract was read.
    • No groundwater Gd concentration was read as a number (the tap water percentages stand in).

    Sources

    The Element Book, gadolinium entry (data/elements/Gd.json and data/reference/text/Gd.json)
    Bau, M. and Dulski, P., Anthropogenic origin of positive gadolinium anomalies in river waters, Earth and Planetary Science Letters 143 (1996) 245 to 255 (abstract only)
    Kulaksiz, S., Rare earth elements as emerging contaminants in the Rhine River, Germany and its tributaries, PhD thesis, Jacobs University Bremen (2012), chapters III to V (the Environment International 2011, Applied Geochemistry 2011 and EPSL 2013 papers)
    Kulaksiz, S. and Bau, M., 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, Earth and Planetary Science Letters 362 (2013) 43 to 50 (read as chapter V of the thesis)
    Verplanck, P. L. et al., Aqueous stability of gadolinium in surface waters receiving sewage treatment plant effluent, Boulder Creek, Colorado, Environmental Science and Technology 39 (2005) 6923 to 6929 (abstract, PubMed 16201612)
    Verplanck, P. L. et al., Evaluating the behavior of gadolinium and other rare earth elements through large metropolitan sewage treatment plants, Environmental Science and Technology 44 (2010) (abstract, PubMed 20397691)
    Lawrence, M. G. et al., Removal of magnetic resonance imaging contrast agents through advanced water treatment plants, Water Science and Technology 61 (2010) 685 to 692 (abstract via Crossref)
    Schmidt, K., Bau, M., Merschel, G. and Tepe, N., Anthropogenic gadolinium in tap water and in tap water-based beverages from fast-food franchises in six major cities in Germany, Science of the Total Environment 687 (2019) 1401 to 1408 (abstract, PubMed 31412473)
    Tepe, N., Romero, M. and Bau, M., High-technology metals as emerging contaminants: strong increase of anthropogenic gadolinium levels in tap water of Berlin, Germany, from 2009 to 2012, Applied Geochemistry 45 (2014) 191 to 197 (abstract as summarised in search results)
    Hatje, V., Bruland, K. W. and Flegal, A. R., Increases in anthropogenic gadolinium anomalies and rare earth element concentrations in San Francisco Bay over a 20 year record, Environmental Science and Technology 50 (2016) 4159 to 4168 (abstract via Europe PMC)
    Brunjes, R. and Hofmann, T., Anthropogenic gadolinium in freshwater and drinking water systems, Water Research 182 (2020) 115966 (abstract)
    Birka, M. et al., Investigating the stability of gadolinium based contrast agents towards UV radiation, Water Research (2016) (abstract, PubMed 26802476)
    US FDA Drug Safety Communication of 19 December 2017: gadolinium-based contrast agents are retained in the body, new class warnings (as reproduced by the Sentinel Initiative)
    Merschel, G. and Bau, M., Rare earth elements in the aragonitic shell of freshwater mussel Corbicula fluminea and the bioavailability of anthropogenic lanthanum, samarium and gadolinium in river water, Science of the Total Environment 533 (2015) 91 to 101 (abstract via Europe PMC)
    Sneller, F. E. C., Kalf, D. F., Weltje, L. and Van Wezel, A. P., Maximum Permissible Concentrations and Negligible Concentrations for Rare Earth Elements (REEs), RIVM report 601501 011 (Bilthoven, 2000), Tables I, II, 2.2 and 4.1, section 5.1 and Appendices 1 to 3
    Rare earth element geochemistry characteristics of seawater and porewater from deep sea in western Pacific, Scientific Reports 7 (2017), Table 1 (Pigafetta basin, 3 to 5663 m)
    Rare-Earth Elements as Natural Tracers for In Situ Remediation of Groundwater (open access, PMC7868090); REE speciation and concentrations at three US groundwater remediation sites
    The occurrence and concentration of rare earth elements in acid mine drainage and treatment byproducts, Part 2: regional survey of northern and central Appalachian coal basins, Mining, Metallurgy and Exploration (OSTI 1577122)
    ISO 17294-2:2023, Water quality. Application of ICP-MS. Part 2: Determination of selected elements including uranium isotopes (element list from the ISO catalogue abstract)
    US EPA, National Primary Drinking Water Regulations (table of regulated contaminants; no lanthanide, rare earth or actinium entry; gross alpha 15 pCi/L)
    Commission Implementing Decision (EU) 2016/902 establishing BAT conclusions for common waste water and waste gas treatment/management systems in the chemical sector (CWW), BAT 12 Tables 1 to 3

    Conventions

    Valence electrons are counted by the usual convention: the outer shell for s- and p-block elements, ns and (n-1)d for the d-block, ns, (n-1)d and (n-2)f for the f-block. Lanthanides and actinides are placed in the f-block with no group number. Electrical conductivity is the reciprocal of the printed resistivity. Ionic radii are Shannon effective radii, six-coordinate unless noted. Where a field reads “not in sources” the value was not found; it is a gap, not a zero. Regulatory limits are the published values and change often, so check the standard in force at your site and the numbers written into your own permit before you design to them.

    Data

    Element records, isotopes, radii and the descriptive text come from PubChem (NCBI), the Los Alamos National Laboratory periodic table, IUPAC CIAAW and the IAEA Atomic Mass Data Center. Appearance, thermal conductivity, electrical resistivity, crystal structure, discovery and the origin of each name come from Wikipedia and Periodic-Table-JSON, used under CC BY-SA 4.0. Hazard classifications come from the ECHA C&L inventory via PubChem. Ionic radii follow R. D. Shannon (1976). The water chapters cite their own sources at the foot of each entry, and are written to the level of Snoeyink and Jenkins, Stumm and Morgan, MWH's Water Treatment and Metcalf and Eddy.