The Great Oxidation Event (GOE), marked by the permanent accumulation of free oxygen in Earth's atmosphere around 2.5 billion years ago, is a pivotal moment in Earth's history. While the rise of oxygenic photosynthesis by cyanobacteria is understood to be the primary driver, the complex interplay between atmospheric oxygenation and the co-evolution of Earth's geosphere and biosphere remains poorly understood. Numerous proxies, including isotopic compositions of carbon and sulfur, trace element concentrations in sediments, and organic biomarkers, have been used to reconstruct past atmospheric oxygen levels. However, these methods often struggle to directly assess the spatially averaged redox state of Earth's crust through time. This study introduces a novel approach using manganese (Mn) minerals as a proxy for crustal oxidation. Manganese, abundant in the Earth's crust and exhibiting variable oxidation states (+2, +3, +4), provides a sensitive indicator of changes in the crustal redox environment. The research question is centered around quantifying the changes in the average oxidation state of crustal Mn occurrences through time and relating this to atmospheric oxygen levels, aiming to understand the time lag between atmospheric changes and their impact on the crust. The significance of this work lies in providing a new, large-scale mineralogical proxy for assessing crustal redox evolution, which has been previously limited by a lack of suitable methods capable of capturing spatially averaged conditions over geological time. This study's findings significantly contribute to a more comprehensive understanding of Earth's complex co-evolutionary processes.

Previous research has utilized various proxies to reconstruct past atmospheric oxygen levels, including carbon and sulfur isotope compositions, trace element concentrations in sediments, and organic biomarkers. These studies have helped establish a general timeline of atmospheric oxygenation, including the GOE and subsequent fluctuations. However, understanding the connection between atmospheric oxygen and the redox state of Earth's crust remains a challenge. Studies like that by Golden et al. (2013) have examined the rhenium content of molybdenite to infer crustal oxidation, revealing a significant time lag between atmospheric oxygen rise and crustal oxidation. Other studies have explored the oxidation of Earth's mantle through hydrothermal processes and subduction. However, comprehensive, spatially averaged data on the evolving redox state of the Earth's crust has remained limited, highlighting the need for a more direct and large-scale approach. This research builds on this existing literature by utilizing the vast dataset of manganese minerals available, offering a new proxy to address the limitations of previous studies.

This study leverages two extensive mineralogical databases: mindat.org and the Mineral Evolution Database (MED). As of November 20, 2015, the researchers collected 22,064 mineral-locality data pairs from mindat.org, of which 2666 had associated geologic ages from the MED. The data included all known Mn-bearing mineral species approved by the International Mineralogical Association, with Mn as an essential element in their chemical formulas. The analysis focused on excluding Mn present as a minor constituent in other minerals, which primarily exists as Mn2+ and would not influence the analysis of higher oxidation states. Similarly, Mn-nodules from the ocean floor were excluded due to their geologically recent origin. The Mn minerals were categorized into three lists based on their Mn oxidation state (+2, +3, +4). Oxidation states were determined primarily from charge balance in the chemical formula. For ambiguous cases, literature on crystal structures, spectroscopic data, and structural refinements were consulted. Minerals containing Mn in multiple oxidation states were included in the corresponding lists. A histogram (Figure 1) displays the frequency of Mn mineral occurrences over time (50 million-year bins), highlighting maxima near supercontinent formation events. To quantify the average oxidation state of Mn through time, the fraction of each oxidation state within each time bin was calculated (Figure 2). Standard errors were determined by fitting the raw Mn mineral counts to the negative binomial distribution. The average oxidation state was calculated as a weighted average of the three oxidation states, and errors were calculated using standard propagation of error. Finally, to quantify the correlation between average Mn oxidation state and atmospheric oxygen reconstructions from previous studies (Berner and Canfield, Bergman et al., Holland), a sinusoidal function was fit to both datasets. This allowed the calculation of an oscillation period and a phase shift representing the time lag between changes in atmospheric oxygen and Mn oxidation state. The oxygen reconstruction from Bergman et al was then shifted forward in time by 66 Myr to highlight the correlation.

The study reveals a strong correlation between the average oxidation state of manganese in Earth's crust and atmospheric oxygen levels over the past billion years. A histogram of dated Mn mineral occurrences (Figure 1) shows maxima coinciding with supercontinent formation events, a well-known trend for other mineral groups. Analyzing the average Mn oxidation state through time (Figure 2), the researchers observed a long-term increase in the proportion of Mn minerals with higher oxidation states (+3 and +4) following the GOE. A noteworthy increase in average Mn oxidation state is clearly visible from 600 Ma to the present, with smaller fluctuations during the Phanerozoic. This trend is particularly evident in the proportions of Mn oxidation states (+2, +3, +4) in the last 600 million years (Figure 2), showing a clear decrease in Mn2+ mineral species relative to Mn3+ and Mn4+ species. The study also shows an increase in Mn oxidation state from +2 to greater than +2.5 during the existence of Kenorland (3.0–2.5 Ga). This early oxidation predates most estimates of the GOE, possibly indicating an earlier form of manganese-oxidizing photosynthesis or the influence of pre-GOE oxygen “whiffs.” A comparison between the average Mn oxidation state and atmospheric oxygen reconstructions (Figure 3) reveals a striking similarity in the patterns of increase and decrease, albeit with a time lag. By fitting both datasets to a sinusoidal function, the researchers determined a time lag of approximately 66 ± 1 million years between changes in atmospheric oxygen and the corresponding changes in Mn oxidation state. This lag is interpreted as the time required for the shallow crust to equilibrate to changes in atmospheric oxygen fugacity through weathering, fracturing, and tectonic processes. The close agreement between the oscillatory periods derived from both datasets (35.7 Ma for oxygen and 35.0 Ma for Mn) further strengthens the correlation.

The observed 66-million-year time lag between changes in atmospheric oxygen and the corresponding changes in the average oxidation state of manganese in Earth's crust is a key finding. This lag is not unexpected, given the limited direct exposure of most crustal materials to the atmosphere. The study interprets this lag as the average time needed for the upper 1 km of the crust to equilibrate to new atmospheric oxygen fugacity through oxidative weathering, fracturing, and tectonic mixing. The difference in time lag between the Mn data and that from the previous molybdenite study suggests that different mineral groups may record changes in crustal oxidation state at different depths, implying a depth-dependent oxidation response. This finding supports the use of different mineral systems to map the crustal oxidation state with depth and time. The strong correlation between atmospheric oxygen levels and the oxidation state of manganese in the crust emphasizes the significant impact of atmospheric oxygen on the redox state of Earth's shallow crust over geological timescales. The study demonstrates the effectiveness of using large mineralogical databases to constrain the evolving redox state of Earth's reservoirs and suggests further analysis of other redox-sensitive metals to refine our understanding of this crucial aspect of Earth's history.

This study presents compelling evidence for the oxidation of Earth's shallow crust over a substantial portion of Earth's history, using manganese minerals as a novel and effective proxy. The 66-million-year time lag between changes in atmospheric oxygen and corresponding changes in Mn oxidation state highlights the timescale involved in crustal equilibration to changing atmospheric conditions. The study suggests that the analysis of large mineralogical databases for other abundant redox-sensitive metals will further refine our understanding of Earth's oxygen fugacity through deep time. Future research could focus on analyzing different mineral groups to determine the timing of oxidation at various depths within the crust, providing a more complete picture of Earth's co-evolutionary processes.

The study acknowledges that preserved Mn minerals represent a range of formation depths and redox states, with a potential bias toward more oxidized Mn minerals due to preferential preservation. The uncertainties associated with the geologic ages of some Mn mineral occurrences could also impact the precision of the time lag calculation. Furthermore, the study's analysis primarily focuses on the shallow crust, and might not fully capture changes in the deeper crust. The interpretation of the early oxidation of Mn predating the GOE remains open to multiple interpretations and requires further investigation.