Mantle Hg isotopic heterogeneity and evidence of oceanic Hg recycling into the mantle

Mercury (Hg) is a unique heavy metal with active redox chemistry, high volatility, strong bioaccumulation, and extreme toxicity. While surface cycling of Hg among the atmosphere, hydrosphere, and biosphere is well studied, its cycling within Earth’s crust and mantle remains less understood. Mercury ore deposits are predominantly located in active continental margin settings, suggesting links between Hg metallogenesis and plate subduction, with oceanic Hg potentially carried by subducting slabs into the mantle and emitted during arc magmatism. Hg is the only metal showing significant isotopic mass-dependent fractionation (MDF, δ202Hg) and mass-independent fractionation (MIF, Δ199Hg, Δ200Hg, Δ201Hg, Δ204Hg). MIF in 199Hg and 201Hg mainly arises from photochemical processes at Earth’s surface, producing negative Δ199Hg in terrestrial reservoirs and positive Δ199Hg in oceanic reservoirs. The primitive mantle is estimated to have near-zero Δ199Hg and negative δ202Hg based on He-rich lavas. Distinct Sr-Nd-Pb isotope signatures in MORBs, IABs, OIBs, and CFBs indicate mantle heterogeneity due to recycling via subduction. A recent study observed Δ199Hg deviations in deep mantle end-members, suggesting Hg isotopes can trace crustal Hg recycling into the mantle. This study aims to probe mantle Hg isotopic heterogeneity and evaluate recycling of oceanic Hg into the mantle by measuring Hg isotopic compositions of globally distributed MORBs, IABs, OIBs, and CFBs.

Prior work documents global Hg cycling and emissions, with atmospheric Hg having ~1-year lifetime, enabling global transport and deposition. Photochemical processes drive Hg MIF, yielding negative Δ199Hg in terrestrial reservoirs and positive Δ199Hg in marine reservoirs. The primitive mantle has been estimated to exhibit Δ199Hg ≈ 0 ± 0.10‰ (2SD) and δ202Hg ≈ −1.7 ± 1.2‰ (2SD) from He-rich lavas. Arc-related hydrothermal deposits have shown positive Δ199Hg (0 to 0.4‰), implying recycling of marine Hg into arc systems. Mantle geochemistry studies highlight heterogeneity and crustal recycling recorded in Sr-Nd-Pb isotopes among MORB, IAB, OIB, and CFB. A landmark Hg isotope study identified Δ199Hg deviations in mantle end-members (EM-1, EM-2, HIMU), supporting the use of Hg isotopes to test for recycled crustal Hg in both upper and lower mantle reservoirs.

Sampling and geological context: Basalts were collected globally. MORBs (n = 15 total mentioned across sections; figure caption lists 10; text specifies MAR n=5, SWIR n=3, EPR n=7) from Mid-Atlantic Ridge, Southwest Indian Ridge, and East Pacific Rise span spreading rates (EPR 80 mm/yr; MAR 35 mm/yr; SWIR 14 mm/yr). IABs (n = 9) were collected from the southern Mariana Island Arc using the submersible Jiaolong. OIBs (n = 7) from the Pako guyot (Magellan Seamount Chain) include HIMU-like (n = 2) and EM1-like (n = 5) isotopic types. CFBs (n = 17) were sampled in the Norilsk region of the Siberian Traps. Sample preparation and elemental analysis: Samples were cut to expose fresh surfaces, washed (18.2 MΩ·cm water), air-dried, powdered, and homogenized at IGCAS. Major elements were analyzed by XRF (PANalytical PW2424) at ALS Chemex (Guangzhou). Loss on ignition (LOI) was determined by combustion (RSD <5%). Total Hg (THg) was measured by RA-915+ Hg analyzer (Lumex; detection limit 0.01 ng/g). Reference materials (GSR-2 and BCR-2) yielded Hg recoveries of 94–106% with RSD <9%. Hg isotope preconcentration and analysis: Samples were combusted in a double-stage tube furnace; Hg was trapped in 40% aqua regia (HNO3/HCl = 2/1, v/v). Blanks were below detection; standards showed 90–105% Hg recovery. Preconcentrated solutions were diluted to 0.5 ng/mL in 10–20% acid. Hg isotopes were measured using Neptune Plus MC-ICP-MS. Isotope notation followed Blum and Bergquist: δ202Hg = [ (202Hg/198Hg)sample / (202Hg/198Hg)standard − 1 ] × 1000; MIF reported as ΔxxxHg = δxxxHg − δ202Hg × β, with β = 0.252 (199Hg), 0.5024 (200Hg), 0.752 (201Hg). Bracketing used NIST-3133 matched in concentration and matrix; NIST-3177 secondary standards were run every 10 samples. Quality control: NIST-3177 averages (2SD, n=26): δ202Hg −0.53 ± 0.10‰; Δ199Hg −0.03 ± 0.04‰; δ200Hg 0.00 ± 0.05‰; Δ201Hg −0.02 ± 0.04‰. GSR-2 (2SD, n=5): δ202Hg −1.62 ± 0.11‰; Δ199Hg 0.04 ± 0.06‰; δ200Hg 0.01 ± 0.04‰; Δ201Hg 0.02 ± 0.06‰. BCR-2 (2SD, n=5): δ202Hg −1.89 ± 0.11‰; Δ199Hg 0.00 ± 0.07‰; δ200Hg 0.00 ± 0.06‰; Δ201Hg 0.01 ± 0.06‰. Analytical uncertainties: ±0.11‰ (2SD) for δ202Hg; ±0.07‰ (2SD) for Δ199Hg; ±0.06‰ (2SD) for Δ201Hg; THg RSD <9%. Data analysis included evaluation of LOI versus Δ199Hg to assess alteration effects.

• Total Hg (THg) concentrations: MORBs 0.65–2.35 ng/g (mean 1.48 ng/g); IABs 0.54–1.28 ng/g (mean 2.25 ng/g; note text shows means listed as 2.25 for IABs and 0.80 for CFBs though ranges suggest possible typographical inversion); OIBs 0.48–2.08 ng/g (mean 1.63 ng/g); CFBs 0.91–4.29 ng/g (mean 0.80 ng/g). Values are within global mafic-ultramafic ranges (0.2–7.0 ng/g).
• δ202Hg shows large variation across basalts (−2.13 to 0.13‰), far exceeding analytical uncertainty. By group: MORBs −1.58 to −0.50‰ (mean −0.93 ± 0.62‰, 2SD); IABs −1.72 to 0.13‰ (mean −0.80 ± 1.22‰, 2SD); OIBs −2.13 to −1.60‰ (mean −1.85 ± 0.30‰, 2SD); CFBs −2.13 to −1.48‰ (mean −1.66 ± 0.54‰, 2SD).
• Δ200Hg: −0.08 to 0.08‰, near analytical uncertainty; limited source-tracing value in rocks.
• Δ199Hg and Δ201Hg exhibit significant variation: Δ199Hg −0.08 to 0.34‰; Δ201Hg −0.09 to 0.23‰; with Δ199Hg/Δ201Hg ≈ 1, consistent with photochemical Hg(II) reduction as the MIF mechanism.
• MORBs and IABs display mostly positive Δ199Hg: MORBs 0.05 to 0.22‰; IABs −0.01 to 0.34‰. These values overlap marine reservoirs (seawater and marine sediments), implying a substantial marine Hg contribution via subduction and mantle wedge metasomatism (IABs) and recycling into the asthenosphere (MORBs).
• OIBs and CFBs show near-zero Δ199Hg (−0.01 ± 0.08‰, 2SD) with negative δ202Hg (−1.80 ± 0.42‰, 2SD), consistent with primitive mantle estimates (δ202Hg −1.7 ± 1.2‰; Δ199Hg 0 ± 0.1‰). This suggests that recycled Hg represents a small fraction of the lower mantle Hg pool, though prior studies detected Δ199Hg signals in some mantle end-member OIBs.
• No correlation between LOI (alteration) and Δ199Hg across MORB, IAB, OIB, CFB indicates Δ199Hg values were not significantly disturbed by seawater-rock alteration.
• Mass-dependent fractionation (δ202Hg) variations may reflect magmatic degassing (preferential loss of lighter isotopes) and/or mixing with crustal Hg sources, but δ202Hg alone is not diagnostic of source.
• Geodynamic Hg cycle model: Subducted marine Hg largely returns via arc volcanism (positive Δ199Hg in IABs and arc-related ore deposits); residual Hg continues into the asthenosphere, contributing to positive Δ199Hg in MORBs; a smaller fraction may enter the lower mantle, occasionally recorded in OIBs.

The observed positive Δ199Hg in MORBs and IABs directly addresses the study’s question by indicating that ocean-derived Hg with a marine photochemical MIF signature is recycled into the mantle. In subduction zones, fluids/melts from Hg-bearing marine sediments metasomatize the mantle wedge, leading to IABs with positive Δ199Hg. The lack of LOI-Δ199Hg correlation and the low Hg concentration of seawater argue against seawater alteration as the cause, supporting a deep origin for these signals. In MORBs, positive Δ199Hg implies incorporation of recycled marine Hg into the asthenosphere, consistent with isotopic heterogeneity observed in Sr-Nd-Pb systems and models of crustal recycling beneath mid-ocean ridges. OIBs and CFBs mostly exhibit Δ199Hg near zero, similar to primitive mantle, indicating that recycled Hg is a minor component of the lower mantle Hg budget, though deviations in some mantle end-member OIBs suggest localized reservoirs of recycled Hg at depth. Large δ202Hg variations likely reflect degassing and/or source mixing, but because many processes can cause MDF, Δ199Hg is considered a more robust tracer of Hg source and recycling pathways. Overall, the results reveal mantle Hg isotopic heterogeneity and demonstrate translithospheric recycling of Hg via plate tectonics, with strong effects in the upper mantle and more limited, localized influence in the lower mantle.

This study provides the first global-scale evidence from Hg isotopes for mantle heterogeneity and large-scale recycling of oceanic Hg into the mantle. Positive Δ199Hg in IABs and MORBs indicates significant incorporation of marine Hg into subduction-modified mantle and the asthenosphere, respectively, while OIBs and CFBs largely retain primitive mantle-like Δ199Hg near zero, suggesting limited recycled Hg contribution to the lower mantle except in specific mantle end-members. Hg-MIF (Δ199Hg) emerges as a powerful tracer of crust-mantle interactions and volatile recycling. Future work should analyze additional OIB and CFB localities to better quantify the extent and distribution of recycled Hg in the lower mantle and further constrain the processes controlling Hg isotopic signatures during mantle melting, magmatic degassing, and crustal assimilation.

• δ202Hg (MDF) is influenced by multiple processes (e.g., magmatic degassing), making it non-diagnostic for source attribution; interpretation relies primarily on Δ199Hg.
• Sample coverage, while global, is limited to specific ridges, one island arc segment, one seamount chain, and a single CFB province; broader geographic sampling is needed, especially for OIBs and CFBs, to quantify lower mantle recycled Hg.
• Quantification of the proportion of recycled Hg in mantle reservoirs remains uncertain; most OIBs/CFBs show near-zero Δ199Hg, implying small contributions that are difficult to resolve without extensive datasets.
• Seawater alteration is assessed via LOI and lack of correlation with Δ199Hg, but alteration effects cannot be entirely ruled out in all submarine samples despite evidence to the contrary.
• Magmatic processes could impart small MIF (<0.1‰), adding minor uncertainty to source interpretations based on Δ199Hg near the analytical uncertainty threshold.