Anomalous fractionation of mercury isotopes in the Late Archean atmosphere

Abstract

Earth's surface underwent a dramatic transition ~2.3 billion years ago when atmospheric oxygen first accumulated during the Great Oxidation Event, but the detailed composition of the reducing early atmosphere is not well known. Here we develop mercury (Hg) stable isotopes as a proxy for paleoatmospheric chemistry and use Hg isotope data from 2.5 billion-year-old sedimentary rocks to examine changes in the Late Archean atmosphere immediately prior to the Great Oxidation Event. These sediments preserve evidence of strong photochemical transformations of mercury in the absence of molecular oxygen. In addition, these geochemical records combined with previously published multi-proxy data support a vital role for methane in Earth's early atmosphere.

Fig. 1. Lithologic and geochemical data for the studied section.

Data are listed in Supplementary Data 1, and include Hg concentrations (ppb), total organic carbon (TOC, wt%), Hg/TOC ratios for samples with TOC > 0.2 wt% (ppb/wt%), Hg isotope values (δ²⁰²Hg, Δ¹⁹⁹Hg, and Δ²⁰⁰Hg, in ‰), and pyrite S-MIF data (Δ³⁶S/Δ³³S; from ref. ). Uncertainties on Hg isotope data correspond to the larger value of either the measurement uncertainty of replicate digests of MESS-2 or the uncertainty of repeated measurements of UM-Almadén. Uncertainty on Δ³⁶S/Δ³³S was computed from the larger of the internal or external uncertainties associated with the raw Δ³³S and Δ³⁶S data (as previously reported). Analytical uncertainties for additional data are encompassed within each individual data point. The gray shaded area indicates the interval where changes in Δ³⁶S/Δ³³S correspond to changes in Hg-MIF values; the vertical purple line illustrates the background Archean Δ³⁶S/Δ³³S ratio; vertical dashed lines also illustrate 0‰ in Δ¹⁹⁹Hg and Δ²⁰⁰Hg.

Fig. 2. Cross plots of geochemical data.

Uncertainties on Hg isotope data correspond to the larger value of either the measurement uncertainty of replicate digests of MESS-2 or the uncertainty of repeated measurements of UM-Almadén; analytical uncertainties for additional data are encompassed within each individual data point. These plots demonstrate a lack of correlation between Hg (ppb) and total sulfur content (S wt%) (a), and generally positive correlations between Hg (ppb) and total organic carbon (TOC, wt%) (b), Δ¹⁹⁹Hg and Δ²⁰¹Hg (both in ‰) (c), and Δ¹⁹⁹Hg and Δ²⁰⁰Hg (both in ‰) (d).

Fig. 3. Temporal plot of mercury isotope data from the rock record.

These include Δ²⁰⁰Hg (in ‰, a) and Δ¹⁹⁹Hg (in ‰, b). Our data are shown as red dots (for Δ²⁰⁰Hg) and blue dots (for Δ¹⁹⁹Hg). Previously published Proterozoic and Phanerozoic data (shown as black dots) are compiled from^–,,,,,,–. The only previously published Hg-MIF data from Archean sediments (shown as gray open circles) were from the Mt McRae shale, which have been controversially associated with transient atmospheric oxygen before the GOE.

Fig. 4. Cross plots of mercury isotope data from marine sediments.

These include δ²⁰²Hg versus ∆¹⁹⁹Hg from late Late Archean sediments (our data) compared with Late Mesoproterozoic shales from Africa, Early Cambrian shales from China, end-Permian sediments from China^,, end-Triassic shelf sediments from North America, and Eocene shales from the Arctic Basin.

Fig. 5. Proposed scenarios for observed mercury isotope fractionation trends.

a Intense photochemical reactions in the absence of an ozone layer lead to the production of large Hg-MIF in the Archean atmosphere. The red color represents volcanic Hg(0), which has been the primary source of Hg to the Earth surface over geologic time. This volcanic Hg(0) is characterized by Δ¹⁹⁹Hg ~0‰ and Δ²⁰⁰Hg ~0‰, and can undergo long-range transport in the atmosphere. Atmospheric Hg(0) was oxidized to gaseous Hg(II), and a fraction of that Hg(II) was photoreduced back to Hg(0). These photochemical processes produced gaseous Hg(0) with negative Δ¹⁹⁹Hg and Δ²⁰⁰Hg values, and gaseous Hg(II) with positive Δ¹⁹⁹Hg and Δ²⁰⁰Hg values. The orange color represents the deposition of gaseous Hg(0) to ocean surface and the reemission to the atmosphere. The light blue color represents the deposition of gaseous Hg(II) through precipitation, the scavenging of seawater Hg (mainly precipitated Hg(II)) by organic particles, and the burial of organic-bound Hg with mainly positive Δ¹⁹⁹Hg and Δ²⁰⁰Hg signals in marine sediments. b In the presence of a hydrocarbon haze, nearly all Hg(0) was oxidized by low molecular weight organic compounds before deposition, scavenging, and burial. The red color represents volcanic Hg(0), which was mixed with background atmospheric Hg(0) carrying negative Δ¹⁹⁹Hg and Δ²⁰⁰Hg values. The orange color represents nearly complete oxidation of Hg(0) to Hg(II), and therefore the deposited Hg(II) is characterized by negative Δ¹⁹⁹Hg and Δ²⁰⁰Hg values. The light blue color represents the mixing of the deposited atmospheric Hg(II) with seawater Hg that is characterized with positive Δ¹⁹⁹Hg and Δ²⁰⁰Hg values. The scavenging and the burial of these two forms of Hg produced marine sediments with slightly negative Δ¹⁹⁹Hg and near zero Δ²⁰⁰Hg signals.