Fingerprinting the Cretaceous-Paleogene boundary impact with Zn isotopes

The Cretaceous–Paleogene (K-Pg) boundary is marked by global geochemical anomalies and mass extinction, widely attributed to a bolide impact at Chicxulub. Prior indicators (e.g., Ir/PGE enrichments, shocked minerals, spherules, Cr and Os isotopic signatures, geophysical and geochronological constraints) strongly support an impact, while the relative role of contemporaneous Deccan volcanism remains debated. However, most geochemical evidence has been phenomenological (addition of exotic material) rather than mechanistic, lacking a process-specific tracer of impact-induced volatilization. Zinc (Zn) isotopes are sensitive to volatility: evaporative loss preferentially removes light isotopes, enriching the residue in heavy Zn (high δ66Zn) while lowering Zn concentration. The study tests the hypothesis that Zn isotope systematics in K-Pg boundary sediments record a volatilization fingerprint produced during impact ejecta formation and transport. To evaluate this, the authors analyze boundary layers spanning proximal to intermediate distances from Chicxulub across varied depositional environments (terrestrial, transitional, shallow to deep marine).

Key prior work established global Ir/PGE anomalies in K-Pg strata and widespread impact ejecta (spherules, Ni-rich spinels, shocked minerals). Chromium isotopes (δ54Cr) indicate a carbonaceous chondrite-like (CM2) projectile; Os isotopes suggest Deccan volcanism largely predated the boundary. Geophysical imaging and geochronology tie the event to Chicxulub. Stable Zn isotopes have proven effective tracers of volatility in tektites, nuclear blast ejecta, and super-heated impact melts, consistently showing Zn loss coupled with heavy isotope enrichment (higher δ66Zn) in residues. In contrast, magmatic differentiation and most terrestrial processes typically fractionate Zn isotopes by ≤~0.1‰, and volcanically driven marine inputs at the Permian–Triassic boundary produced increased Zn concentrations with decreased δ66Zn, opposite to expected volatilization trends. Secondary alteration in soils/sediments generally yields small, inconsistent Zn isotope shifts (≤~0.3‰ among products/reactants). These foundations motivate applying Zn isotopes to fingerprint impact-related volatilization in the K-Pg boundary.

Sampling targeted K-Pg boundary clays and adjacent strata at five locations spanning depositional settings and distances from Chicxulub: terrestrial (Montana), transitional (Missouri), shallow marine (Mississippi and ODP Leg 165 in the Caribbean), and deeper marine (New Jersey; ODP Leg 174AX Bass River). All sections contain hallmark K-Pg ejecta (type 1 and 2 spherules), ash layers, and reported Ir/PGE anomalies. Four to five samples were collected within boundary clays and within ~40 cm above/below at Mississippi, ODP Leg 165 (sites 1001A/B), and New Jersey; single boundary samples were obtained in Montana and Missouri, with broader stratigraphic context sampled there. Visible type 1 spherules from Mississippi were hand-picked for phase-specific analysis. A basal limestone from ODP Leg 165 (site 1001) was selected to represent pre-impact carbonate platform composition. Approximately 1–8 g of rock was powdered (<63 μm). For isotopes, ~200–300 mg was digested in two stages: (1) 10 M HF + 8 M HNO3 (6 ml), heated and dried; (2) aqua regia, with complete digestion visually confirmed. Zinc was purified by anion exchange (BioRad MP-1, 200–400 mesh, HCl form) following established protocols, with two passes to remove matrix. Zn isotopes were measured by MC-ICP-MS (Thermo Neptune Plus) at Rutgers and Penn State under low resolution conditions. Solutions (~200 ppb Zn) were doped with 100 ppb Cu (NIST 976) to correct mass bias using the exponential law, and standard–sample–standard bracketing employed AA-ETH Zn; data are reported relative to the (now unavailable) JCM Lyon standard by applying a +0.27‰ offset. External reproducibility across four sessions is ±0.05‰; duplicates and procedural standards (USGS BVHO-2: +0.28 ± 0.03‰, n=8) confirm precision. Zn concentrations were measured by quadrupole ICP-MS on 100 mg aliquots, gravimetrically calibrated; BVHO-2 concentrations were within 8% of reported values. SEM imaging (JEOL 6460) documented spherule textures. Modeling: Binary mixing between a high-δ66Zn, low-Zn volatilized end-member (modeled from average tektites; ~1 ppm Zn, δ66Zn ~+1.8‰) and background sediment end-members (carbonate ~8 ppm Zn, δ66Zn ~+0.37‰; terrestrial ~70 ppm Zn, δ66Zn ~+0.27‰) evaluated mixing trends. Rayleigh distillation models compared observed δ66Zn vs normalized Zn concentration with empirical fractionation factors: tektites α≈0.999; K-Pg data encompassed by 0.99975<α<0.99995, reflecting evaporation under non-vacuum conditions and mixing attenuation.

• Boundary layer sediments systematically show elevated δ66Zn and decreased Zn concentrations relative to adjacent strata and to typical continental crust/carbonates. Reported increases in δ66Zn within boundary layers are approximately: +0.7‰ at the Caribbean site, +0.3‰ at Missouri and Montana, and +0.1‰ at Mississippi and New Jersey.
• Tabled examples include: ODP Leg 165 boundary samples with δ66Zn(Lyon) up to +1.10‰ (Zn ~165 ppm) and a very light outlier at the top of the boundary (−0.56‰, Zn ~13 ppm); Montana K/Pg boundary clay δ66Zn(Lyon) ~+0.67‰ with Zn ~17 ppm; Mississippi hand-picked spherules δ66Zn(Lyon) ~+0.39–0.43‰ with Zn ~357–384 ppm.
• Alternative sources/processes are inconsistent with the observed pattern: magmatic differentiation and melting of target rocks generate ≤~0.1‰ Zn isotope fractionation; volcanism (e.g., Deccan) would increase Zn concentration while lowering δ66Zn (as seen near the P–T boundary), opposite to observations; inheritance from target lithologies or meteoritic material is implausible given typical Zn concentrations and isotope ranges of crust and carbonaceous chondrites; secondary alteration/redox processes produce smaller, non-systematic fractionations (<~0.3‰) and stratigraphic patterns argue against leaching.
• The most parsimonious explanation is partial Zn volatilization during impact ejecta formation/transport, enriching residues (spherules, ejecta) in heavy Zn and lowering Zn concentrations. δ66Zn correlates inversely with Zn concentration in a manner consistent with Rayleigh-type evaporation.
• Binary mixing models between volatilized ejecta and background sediments reproduce two trends (carbonate vs terrestrial/transitional backgrounds). In transitional/terrestrial settings, roughly ~10% of Zn is derived from partially volatilized ejecta; marine carbonate settings show larger apparent fractions due to lower background Zn, enhancing the signal.
• Empirical fractionation factors indicate evaporation under pressured conditions with diffusion/back-condensation effects: tektites α≈0.999; K-Pg data fall within 0.99975<α<0.99995, markedly less fractionation than pure vacuum Rayleigh (α≈0.985), and further attenuated by mixing with endogenous Zn (~10:1 dilution).
• Marine sites (Mississippi, New Jersey) exhibit attenuated excursions (often within continental crust δ66Zn uncertainty) because higher authigenic Zn concentrations dilute the ejecta signal; variations within boundary clays imply syndepositional mixing driven by mass flows/tsunami/turbulence.
• Mass balance implies a complementary isotopically light Zn reservoir; a very light ODP Leg 165 sample at the top of the boundary (δ66Zn(Lyon) −0.56‰) may represent later fallout/back-condensation of light Zn, analogous to some tektites with light δ66Zn.

The data provide a mechanistic tracer—heavy Zn isotope enrichment coupled with decreased Zn concentrations—unequivocally consistent with impact-driven volatilization. This directly links a specific process (evaporation during ejecta formation and transport) to the K-Pg boundary layer geochemistry, complementing prior phenomenological indicators. Mixing models show plausible contributions (~10% of Zn) from volatilized ejecta in terrestrial/transitional settings and higher apparent fractions in carbonate settings due to low background Zn, aligning with depositional context and documented ejecta presence. The empirical fractionation factors and adherence to Rayleigh-type trends, albeit attenuated, indicate evaporation under non-vacuum conditions with diffusion and back-condensation moderating isotopic effects and subsequent dilution by endogenous sedimentary Zn. The attenuated and variable signature in marine sections suggests syndepositional mixing processes (e.g., tsunami/mass flows) influencing Zn budgets across clay subunits. Mass balance considerations predict complementary light Zn deposited later within boundary sequences; the observed very light δ66Zn at the top of the Caribbean boundary may represent this component. Overall, Zn isotopes robustly fingerprint volatilization at the K-Pg, advancing from circumstantial to process-based evidence for impact ejecta contributions in boundary sediments.

This study demonstrates that K-Pg boundary sediments bear a distinct Zn isotopic fingerprint of impact-induced volatilization: elevated δ66Zn correlated with reduced Zn concentrations. Alternative explanations (magmatism, volcanism, target or meteoritic inheritance, secondary alteration) cannot account for the observed patterns across diverse depositional settings. Binary mixing and Rayleigh-type models indicate contributions of partially volatilized ejecta to boundary sediments (order ~10% of Zn in terrestrial/transitional settings) and reveal evaporation under non-vacuum conditions with diffusion/back-condensation effects. These results provide a mechanistic geochemical tracer for bolide impacts archived in sedimentary layers. Future work should target: (1) high-resolution sampling within and above boundary clays to locate complementary isotopically light Zn reservoirs; (2) phase-specific analyses of fallout products (spherules, glasses) across multiple sites; (3) broader application of Zn isotopes to other impact horizons to evaluate the generality of this fingerprint; and (4) improved constraints on end-member compositions and fractionation parameters under variable P–T–fO2 conditions.

• Sampling density varies by site; only single boundary samples were available for some locations (Montana, Missouri), limiting statistical robustness.
• Marine sites show attenuated signals often within crustal δ66Zn uncertainty due to higher background Zn, complicating mixing interpretations.
• Binary mixing uses tektites as a proxy for volatilized ejecta; actual ejecta composition and Zn content/isotopic range are variable, introducing model uncertainty.
• Fractionation modeling is qualitative regarding local P–T conditions, extent of volatilization, diffusion, and back-condensation; α values are empirical envelopes rather than uniquely constrained.
• It is not currently possible to deconvolve Zn contributions volatilized from target carbonates versus basement gneiss versus the bolide.
• The complementary isotopically light Zn reservoir is inferred by mass balance but not comprehensively located or characterized.
• Some sample labels/sites have limited ancillary data (e.g., ash/spherule abundance quantification), and carbonate baseline includes minor ash contamination.