When did oxygen flood into the Earth's atmosphere? Credit: NASA

How quickly did oxygen build up in Earth's early atmosphere? An analysis using chromium isotopes trapped in ancient ocean deposits has now provided an unexpected insight into this longstanding puzzle.

The usual geochemical tale is a one-two punch of oxygen molecules flooding the atmosphere, first around 2.45-2.2 billion years ago in what geologists call the Great Oxidation Event (GOE), followed by another steep rise around 750 million years ago. But details have been elusive. Attempts to use molybdenum, rhenium and other metal isotopes to understand oxygen's earlier rise have given a range of results about when the oxygenation began, how quickly it progressed and whether it was continuous. That leaves the rise of oxygen at the GOE as "one of the two or three big questions about the early Earth", says Tim Lyons of the University of California, Riverside.

The latest finding comes from a team led by Robert Frei of the University of Copenhagen, Denmark, who sampled banded iron formations — an iron-rich sedimentary rock — dating from around and in between the two main periods of intense oxygen increases. They show that oxygen snuck into surface ocean waters 2.8–2.6 billion years ago — at least 200 million years earlier than predictions based on analyses of other metal isotopes.

More surprisingly, they also claim that around 1.9 billion years ago, oxygen levels actually dipped back down to almost where they were before the GOE, at less than 1% of today's levels. That's "the most interesting part of the story", says Don Canfield of the University of Southern Denmark in Odense and part of the research team who publish their conclusions in Nature1.

Heavy stuff

The team's method relies on how chromium responds to changing levels of oxygen in the air. Without much oxygen in the atmosphere, chromium is locked into rocks in a form where each atom has three fewer electrons than elemental chromium metal — a reduced, or +3, oxidation state.

Were there pulses of life associated with these pulses of oxygen? Robert Frei , University of Copenhagen, Denmark

But as oxygen levels rise, metallic manganese in the same rocks is converted into manganese oxide, which then steals electrons from the landlocked chromium. The result is an oxidized +6 form of chromium which is much more likely to be dissolved by rainwater and washed into the ocean. Once there, chromium reacts with iron metal and is incorporated into the banded iron formations in its +3 form.

Crucially, the heavy chromium isotope, chromium-53, is more likely to be oxidized and washed into the ocean than its lighter cousin chromium-52. This means that measuring the relative amounts of heavier and lighter chromium isotopes in the banded iron formations indicates how much oxygen was in the atmosphere at the time they were locked into the rocks.

"The technique is terrific," says Robert Hazen of the Geophysical Laboratory, the Carnegie Institution for Science, Washington DC, adding that it seems more sensitive to oxygen shifts than other metal proxies. But he adds that the chromium ions may be reacting with more than just manganese oxide and iron, which may make the picture even more complex. "What we need to understand much better is the weathering process of chromium," agrees Canfield.

Lyons, who did not work on the research, says he is particularly pleased to see that the new proxy continues to work long after the GOE, when the signals of other metal isotopes get swamped by the growing amounts of oxygen present. He adds that wavering oxygen levels, rather than a simple upward trend, has big implications for the development of life at the time.

"Were there pulses of life associated with these pulses of oxygen?" wonders Frei, who hopes that the new findings will foster discussion about whether life forms were photosynthesizing long before the GOE to create these early "whiffs" of oxygen2,3. Whether primitive life was responsible for giving the Earth's atmosphere its first taste of oxygen is "a whole different issue that needs to be investigated", says Frei. For now, he adds, both the new chromium isotope method and the team's findings need to be confirmed by testing more rocks.