Radiation damage allows identification of truly inherited zircon

Inherited zircon older than their host rocks are widely reported in felsic, mafic, and ultramafic rocks across oceanic and continental settings. While zircon can be recycled through the mantle and preserve U-Pb ages under certain conditions, experiments indicate rapid zircon dissolution in mafic melts, raising questions about the origin of old zircon in such rocks. During a study of a Pliocene mid-ocean ridge gabbro from the Mohns Ridge, grains with Precambrian ages were recovered alongside expected young grains, implying possible mantle recycling or contamination. Because contamination during sample handling is a persistent concern—especially with low zircon yields—there is a need for a rapid, non-destructive method to verify the true origin of zircon grains. Raman spectroscopy can characterize radiation damage at micrometer scale, and laboratory studies show that radiation damage in zircon anneals at >700–900 °C on hours-to-days timescales; thus, zircon incorporated into a melt should exhibit crystallinity comparable to co-crystallized magmatic zircon even if older U-Pb ages are retained. This study tests whether radiation damage metrics can distinguish contaminated from truly inherited zircon by combining single-grain U-Pb geochronology with Raman spectroscopy on: (1) a Pliocene mid-ocean ridge gabbro (Case 1) and (2) Cenozoic dacitic–trachydacitic sills from the opening of the North Atlantic (Case 2).

Prior work documents inherited zircon in oceanic and continental settings (e.g., Mid-Atlantic Ridge, ocean islands, island arcs). Some studies propose that zircon can survive mantle residence and preserve U-Pb ages, implying recycling of continental material into the mantle. However, experiments demonstrate rapid zircon dissolution in mafic melts, challenging survivability. Low-temperature thermochronometers (fission tracks, zircon/apatite (U-Th)/He) can constrain recent thermal histories but are destructive and sample-quality dependent. Raman spectroscopy has been used extensively to quantify radiation damage (metamictization) and its annealing kinetics in zircon; partially damaged zircon can recover crystallinity at high temperatures over short durations. Calibrations exist that relate Raman ν3(SiO4) peak position and FWHM to α-dose and damage state, enabling calculation of radiation damage ages. These frameworks suggest that zircon recrystallized or heated in magma should exhibit annealed Raman signatures similar to magmatic zircon, even if inherited cores retain older U-Pb ages.

Study design: Two case studies were analyzed. Case 1 targets zircon from a Pliocene mid-ocean ridge gabbro (Mohns Ridge) expected to crystallize at 3–5 Ma. Case 2 targets zircon from Cenozoic dacitic–trachydacitic sills (Gjallar Ridge) intruding Cretaceous sediments during North Atlantic opening, where inherited continental zircon is likely.

Sampling and petrography: Case 1 gabbro sample GS16A-ROV6-016 was collected by ROV in 2016 at 2449 m bsl at Schulz Massif (73.7298 N, 7.5191 E). Zircon grains are subhedral–euhedral (length 53–230 μm) with oscillatory/sector zoning. Case 2 samples 13C-ROV6-5 (dacite; 65.8958 N, 1.7944 E; 2205 m bsl) and 13C-ROV7-1 (trachydacite; 65.8979 N, 1.7846 E; 2755 m bsl) were collected in 2013; zircon occurs in phenocrysts and matrix, euhedral–subhedral (61–820 μm), commonly with core–rim textures.

Mineral separation and imaging: ~10 kg of gabbro and 2–3 kg of dacite/trachydacite were crushed, processed on a Wilfley table, magnetic fractions removed (Frantz), and zircon concentrated with diiodomethane heavy liquid. Grains were hand-picked, mounted in 25 mm epoxy blocks, and polished. Imaging used transmitted/reflected light microscopy and SEM (BSE and CL) to document internal zoning and morphology; grain metrics were measured in ImageJ.

U-Pb geochronology: LA-ICP-MS at Bergen Geoanalytical Facility employed a 193 nm excimer laser (RESOlution M-50 LR) coupled to a Nu Attom ES HR-SC-ICP-MS. Ablation: 26 μm spot, 5 Hz, 2–2.5 J/cm², 30 s ablation after 15 s baseline. Masses measured: 202,204Hg, 206,207,208Pb, 232Th, 235U, 238U (with 238U calculated from 235U in attenuated mode). Data reduction in Iolite 4 (VizualAge U-ComPbine) corrected blanks, mass bias, and downhole fractionation; ages normalized to zircon 91500; QC standards GJ-1, Plešovice, Mud Tank were analyzed. Analyses with >2% uncertainty or discordance were excluded from weighted mean calculations. IsoplotR was used for weighted averages and KDEs.

Raman spectroscopy and radiation damage quantification: Raman spectra were acquired with a Labram-HR system (514 nm Ar-ion laser, ~7–10 mW at sample, 50× objective, 1800 gr/mm grating) and calibrated to the 520.7 cm−1 Si line. Background was subtracted and peaks fitted in Fityk using Pseudo-Voigt functions. The ν3(SiO4) band near 1008 cm−1 was used to determine peak position and FWHM; instrumental broadening was corrected using Váczi’s empirical formula. Radiation damage was assessed via FWHM and peak position shifts, which increase and decrease respectively with accumulated α-dose.

Dose and annealing modeling: Internal α-dose D was computed from U and Th concentrations and U-Pb ages using decay constants for 238U, 235U, 232Th. FWHM–dose relations used calibration curves for unannealed zircon (after Nasdala) and empirical models relating FWHM to dose (Palenik; refined by Váczi): FWHM = A1 − A2 exp(−B2 D), with parameters from the literature. Radiation damage ages and expected FWHM evolution through time were modeled for a range of effective U (eU) values to compare measured FWHM of young and old domains. Assumptions include representativeness of spot analyses for grain domains, potential zoning, and spot-size disparities between LA-ICP-MS (~26 μm) and Raman (~2–3 μm).

Case 1 (Pliocene mid-ocean ridge gabbro, GS16A-ROV6-016):

• Out of 44 recovered grains, 18 were selected for U-Pb; one had insufficient Pb. Seven grains yield Pliocene U-Pb with weighted mean 206Pb/238U age 4.15 ± 0.13 Ma (MSWD = 1.2, n = 7). These grains have low U (8–64 ppm) and Th (7–72 ppm); eU = 10–81 ppm.
• Young magmatic grains are subhedral–euhedral, oscillatory/sector zoned, with mean ν3 peak position 1008.8 cm−1 (range 1008.0–1009.6) and corrected FWHM mean 3.1 (range 2.6–3.5), consistent with well-ordered structures.
• Ten grains show Precambrian U-Pb dates, 2724–554 Ma, with U = 82–541 ppm and Th = 53–396 ppm. Their ν3 positions span 1007.8 to 998.0 cm−1 and FWHM 4.4–14.9, indicating substantially elevated radiation damage.
• Calculated internal doses for these old grains plot inconsistently relative to expectations for grains annealed at ~5 Ma; collectively, the Raman signatures show much greater damage than annealed zircon would display. Conclusion: the Precambrian grains are contaminants introduced during sample processing, not truly inherited zircon.

Case 2 (Cenozoic dacitic–trachydacitic sills, Gjallar Ridge; 13C-ROV6-5 and 13C-ROV7-1):

• Weighted mean 206Pb/238U ages: 54.21 ± 0.13 Ma (MSWD = 2.16; n = 46) and 55.57 ± 0.12 Ma (MSWD = 4.14; n = 63), consistent with regional TIMS ages for NE Atlantic sills.
• Seventy-nine analyses from rims and cores yield older dates between 80 and 2200 Ma. U concentrations: mean 187 ppm (median 151; range 12–1430 ppm); Th 5–759 ppm; U/Th mean 3.4 (median 3.3); eU 55–857 ppm (mean 203; median 172).
• Raman on 87 dated grains (116 measurements, including 37 rim–core pairs): grains are well ordered with ν3 peak mean 1008.2 cm−1 (range 1006.5–1009.2) and corrected FWHM mean 4.3 (median 4.2; range 3.2–6.0). Cores generally have slightly lower ν3 positions and higher eU than rims.
• Inherited (older-than-host) cores and young rims cluster together in FWHM–peak position space, indicating similar crystallinity and shared recent thermal history; most old cores plot below the unannealed calibration in FWHM–dose space, evidencing post-formation annealing in the magma.
• Some young grains plot above unannealed trajectories, implying additional internal tensional strain due to substitution or rapid growth. The observed FWHM range matches modeled radiation damage accrued since ~55 Ma for eU ≈ 200–1500 ppm.

Overall: Raman-detected radiation damage distinguishes contaminated old zircon (case 1) from genuinely inherited zircon that experienced magmatic annealing (case 2).

The study addresses whether radiation damage metrics can verify the origin of old zircon grains found with young magmatic assemblages. In the mid-ocean ridge gabbro (case 1), Precambrian grains exhibit high disorder (broad FWHM, lowered ν3 positions) incompatible with annealing at ~5 Ma, pointing to laboratory contamination rather than true inheritance. Conversely, in the Cenozoic sills (case 2), older cores have crystallinity indistinguishable from ~55 Ma magmatic rims and young grains, demonstrating annealing during sill emplacement and supporting genuine inheritance. These results provide a robust framework to discriminate contamination from inheritance using rapid, non-destructive Raman spectroscopy combined with U-Pb and trace element data. The approach is especially powerful in young magmatic systems where the time contrast is large. Application to oceanic settings can refine interpretations of crustal recycling, potentially overturning some prior claims of inherited zircon where contamination was not excluded. The absence of truly inherited zircon at the Mohns Ridge locale suggests either spatial or temporal variability in continental recycling beneath ridges, or that prior reports elsewhere may involve contamination; more targeted, in situ studies are needed. The method also enables re-evaluation of archived mounts to verify debated hypotheses about concealed microcontinents or mantle-derived continental material.

This work demonstrates that Raman spectroscopy of zircon radiation damage, integrated with single-grain U-Pb geochronology, can distinguish contaminated from truly inherited zircon. In a Pliocene mid-ocean ridge gabbro, Precambrian grains are identified as contaminants due to excessive radiation damage inconsistent with annealing at ~5 Ma. In Cenozoic dacitic–trachydacitic sills, older cores share annealed crystallinity with ~55 Ma magmatic zircon, confirming genuine inheritance and a shared thermal history. The two-step approach offers a rapid, non-destructive tool to validate inherited zircon across diverse geologic settings, improving constraints on mantle–crust recycling and informing geodynamic models. Future work should systematically calibrate compositional/structural effects on Raman bandwidth, quantify time–temperature dependencies across broader conditions, and apply the method to re-examine previous claims of inherited zircon—particularly in oceanic environments—to refine the extent and mechanisms of continental material recycling.

• Annealing depends on both temperature and duration; low-temperature or very rapid cooling events may yield negligible annealing, complicating interpretation in some tectonic contexts (e.g., lower crustal residence, kimberlites).
• Raman FWHM–dose calibrations have uncertainties and potential intercept offsets due to compositional/structural effects; internal strain from substitution or rapid growth can elevate FWHM independent of α-dose.
• Assumptions that LA-ICP-MS (∼26 μm) and Raman (∼2–3 μm) spot analyses represent the same domain can be violated by fine-scale zoning, leading to mixed ages or mismatched dose estimates.
• At low eU, FWHM evolves slowly with time; combined uncertainties in U–Th concentrations and Raman parameters may render radiation damage ages imprecise for low-eU grains.
• Potential contamination during sample preparation can never be entirely excluded; careful lab protocols and cross-checks remain essential.