Consider the antique indignity of it: some of the oldest rocks in this story were already more than a billion years old when oxygen deigned to show up in Earth’s atmosphere. They are lavas, cooled 3.48 billion years ago in what is now South Africa, and they still carry a chemical memory — faint, but legible — of the hot rock far below the crust where they first melted. Now a team led by Li-Juan Xu, publishing in Nature Communications, argues that memory solves, or at least sharpens, one of the great open questions of planetary history: why the Great Oxidation Event happened when it did.
The event itself — oxygen’s first great gathering in the air, between about 2.4 and 2.3 billion years ago — was, as the study puts it, “a critical step in establishing Earth’s long-term habitability.” Living things made the oxygen. But volcanic gases kept consuming it, and one leading idea says oxygen only began to pile up when that appetite shrank. Xu’s team argues the appetite shrank because of what was happening far below: between roughly 3.5 and 1.9 billion years ago, the mantle itself grew more oxidized — a little more like rust — and volcanoes fed by that altered rock breathed out gas that ate less oxygen.
Three Camps, One Argument
Under the thin crust lies the mantle, the thick layer of hot rock that creeps along almost imperceptibly. When part of it melts, the liquid rises and erupts, carrying chemical traces of where it came from — each old lava flow effectively a letter posted from the deep planet. Whether that deep rock’s oxidation state has changed over time has split researchers into three camps. Robert Nicklas and colleagues say the mantle grew steadily more oxidized. Dante Canil’s vanadium measurements say its oxygen level was fixed more than 4 billion years ago, back when the young planet’s surface was an ocean of molten rock, and never budged. Fangyi Zhang and colleagues counter that ancient lavas only look less oxidized because they melted deeper, under heavier pressure.
Three answers that cannot all be right suggest the trouble may lie, as the study puts it, “not only in the geological record itself, but potentially in the proxies used to decode it” — the chemical clues deployed to read old rock. The favourite clues, metal ratios like vanadium’s, get smudged by how the rock melted, by crust mixing in, by weathering long afterward.
Heavier Marbles, Hungrier Rock
Enter chromium. It comes in slightly different weights — isotopes — that behave almost identically, and fine instruments can count the heavy against the light. Inside hot lava, chromium takes two main forms: one favoured when oxygen is scarce, the other when oxygen is plentiful. The oxygen-rich form binds its neighbours more strongly, and stronger bonds favour the heavier atoms — so the heavy-to-light balance swings whenever the oxygen supply does. Crucially, chromium answers to different influences than vanadium, making it an independent witness.
The team assembled 66 samples of komatiite and a close relative from nine lava flows aged 3.48 to 1.87 billion years, drawn from South Africa, Zimbabwe, Canada and the old bedrock of northern Europe, with Igor Puchtel helping to collect them. Komatiites are the deep-divers of the lava world: they form when a third to a half of a chunk of mantle melts at once. Most samples had barely been touched by water since cooling, and their chromium showed no tie to signs of weathering. In a Beijing lab, Xu — who conceived the study — dissolved powdered rock, separated the chromium and weighed the atoms.
The result, stark in its simplicity: across 41 older lavas (3.48 to 2.69 billion years old), chromium ran slightly heavy and scattered widely; the 25 younger lavas (2.41 to 1.87 billion years old) clustered tightly on a value “indistinguishable from the modern mantle value.” Heavier chromium points to less oxygen at melting, because under oxygen-poor conditions the light atoms linger in the solid left behind while the heavy ones ride the escaping melt. Working backward to the source rock, the team estimated the older lavas’ mantle had roughly a tenth as much available oxygen as the younger ones’ — a rough, model-dependent figure, the authors caution, but one that kept reappearing.
Cross-checks piled up like cordwood. Chromium-rich mineral crystals from komatiites worldwide held slightly more rust-like iron in the younger lavas — an independent oxygen signal. And the team showed that a 400-degree Celsius temperature error, or melting at very different depths, would shift the chromium by about the instruments’ margin of error — far too little to fake the pattern.
Sunken Seafloor, Rising Sky
Why did the mantle change? Cooling played a part — the models put the older mantle about 100 degrees Celsius hotter, pushing melting deeper and producing less oxidized lava. But tracking the share of iron in its rust-like form (a measure pressure and heat leave alone) told the bigger story: from about two of every hundred iron atoms in the older mantle to as many as three in the younger, still short of the roughly four estimated for today. Heat alone couldn’t match some younger lavas. The main driver, the authors write, “was not a physical change, but a chemical one” — chemistry delivered from above. Ocean floor soaked in seawater picks up rust-like iron, and when that seafloor sinks back into the Earth at the plate margins — subduction — it carries oxidation down with it, a route Lei Gao, Peter Cawood and colleagues had already tied to the older mantle.
The payoff for the sky: deep, oxygen-poor melting sends carbon and hydrogen up as methane and hydrogen gas, both ravenous for oxygen; shallower, more oxidized melting sends up carbon dioxide and water vapour, already sated. As the mantle cooled and swallowed oxidized seafloor, both shifts pushed volcanic gas toward consuming less of what life was making — a lasting drop that, the authors propose, helped set off the Great Oxidation Event. Though, they add, “establishing its dominance will require future quantitative mass balance modeling” — and the lava record skips from 2.69 to 2.41 billion years ago, leaving the exact sequence above and below unresolved. Whether oxygen leaking from the core contributed, they found too little data to judge. What remains is the narrow, strange certainty: the rock beneath the crust kept changing long after the planet’s youth, and two layers of the Earth that never touched were keeping each other’s accounts.