The great oxygenation of the Earth's atmosphere revisited

22 June 2018

To understand when, how, and how quickly oxygen has become a component of our atmosphere between about 2.5 and 2.2 billion years ago, an international research team involving RSES-ANU has studied the systematics of the four isotopes of sulfur in more than 700 meters of Australian sedimentary deposits. The results show that the oxygenation of the planet began much earlier than traditionally admitted and that its recording was not synchronous from one continent to another (Australia, South Africa, North America) but spread in time over almost 300 million years. This apparent shift reflects a local effect related to oxidative weathering of older continental surfaces.

In the absence of oxygen in the atmosphere, the UV photolysis of sulfur dioxide (SO2) released by volcanic activity results in the production of sulfur compounds characterized by very specific isotopic fractionations defined as “mass independent” (noted , MIF-S). When dissolved in the ocean, these sulfur compounds transfer this isotopic anomaly to the sedimentary record during their precipitation in the form of pyrite, for example. In the presence of atmospheric oxygen, these particular mass-independent isotopic fractionations disappear. The great oxygenation of the Earth's atmosphere (Great Oxidation Event, GOE) between 2.5 and 2.2 billion years ago (Ga) was defined as the time interval during which a sufficient amount of atmospheric oxygen was present to prevent the production and transfer of these isotopic anomalies into the sedimentary record. The disappearance of these isotopic anomalies in South African sediments over a few meters of sediment thickness, led previous studies to propose that the increase of oxygen in the atmosphere was rapid (less than 10 million years) and globally synchronous at around 2.32 Ga worldwide. However, the presence of large sedimentary gaps in the South African sequences implies that this oxygenation model remains loosely constrained.

“In order to better constrain the mechanisms, magnitude, and duration of the GOE, we conducted a drilling campaign in the Hamersley Basin in Western Australia to study a representative sampling that intersects the period between 2.5 and 2.2 Ga associated with the GOE. Unlike its equivalents in South Africa and North America, the sedimentary sequence studied, the Turee Creek Group, does not show major sedimentary discontinuities,” said Professor Pascal Philippot from University of Montpellier and IPGP, lead investigator of the study.

The analysis of isotopes of sulfur with high stratigraphic resolution shows a relatively homogeneous MIF-S signal of low amplitude (1 ± 0.5 ‰) on all the cores. This signal is punctuated by several sedimentary intervals in which the sulphides do not exhibit MIF-S anomalies. The presence of deposits without MIF-S implies that a significant amount of oxygen were present in the atmosphere as early as 2.45 Ga. The MIF-S signal on the order of 1 ‰ represents the average of the isotopic anomalies measured in the sulphides of the Archean period (4.0 to 2.5 Ga) prior to the GOE. The record of such an anomaly over more than 700 meters of drill cores cannot be explained by atmospheric processes, but it is best attributed to oxidative weathering of older (Archean) continental surfaces and the recycling of a sulphate reservoir of homogeneous isotopic composition of the order of 1‰ in the ocean. “This model allows us to explain that the MIF-S record in sediments of South Africa, North America, and Australia is not synchronous because it depends on local weathering surfaces. These results imply that the current paradigm of defining the GOE at 2.33-2.32 Ga based on the last occurrence of MIF-S in South Africa must be abandoned”, said Professor Pascal Philippot.

RSES-ANU scientists Dr Janaína Ávila and Professor Trevor Ireland, co-authors of the research published in Nature Communications, said the detailed and extensive dataset produced in this study has enable to constraint the timing of the Great Oxidation Event more accurately than previous studies. “Our data indicate that since ca. 2.45 Ga free oxygen was an important component of the Earth’s atmosphere being capable to drive oxidative weathering on land”, said Dr Janaína Ávila.

“Before 2.45Ga, the Earth’s atmosphere and oceans had extremely low levels of oxygen. Several studies indicate that the O2 content of the atmosphere was probably less than 0.001% of the present atmospheric level. During the GOE, the transition from a low oxygen atmosphere to an atmosphere characterized by oxic-suboxic conditions is linked in time with a series of global glacial events. Our data suggest that the rise of atmospheric oxygen occurred around 2.45 Ga or earlier and was not a rapid and synchronous event from one continent to another as previously thought”, said Dr Janaína Ávila.

“Due to recent advances on instrumental capabilities of SHRIMP-SI, an ion microprobe developed by the ANU-Research School of Earth Sciences, we now have the ability to measure with high precision the four isotopes of sulfur in individual domains in single grains. Ion microprobe analysis offers spatial control of measurements that is unmatched by any other technique, allowing isotopic data to be correlated with other geochronological, geochemical, and textural information”, said Dr Janaína Ávila.

Laboratories and organizations involved: Paris Institute of Earth Physics (IPGP / CNRS / Paris Diderot University) and Montpellier Geosciences (Montpellier Geosciences / OREME, University of Montpellier / CNRS / West Indies University), Research School of Earth Sciences (Australian National University, Australia), Ocean Geosciences Laboratory (LGO / IUEM, CNRS / UBO / UBS), John de Laeter Center for Isotope Research (Curtin University, Australia), Biogeosciences (EPHE / University of Bourgogne Franche-Comté / CNRS) and School of Biological, Earth and Environmental Sciences (University of New South Wales, Australia).

Source: Globally asynchronous sulphur isotope signals require re-definition of the Great Oxidation Event. 2018. Philippot, P., Ávila, J., Killingsworth, B., Tessalina, S., Baton, F., Caquineau, T., Muller, E., Pecoits, E., Cartigny, P., Lalonde, S., Ireland, T., Thomazo, C., Van Kranendonk, M.J. and Busigny, V., Nature Communications, DOI: 10.1038/s41467-018-04621-x, 8 juin 2018