A rock picked up 14 metres from a lunar module in 1972 and a lump of grey feldspar chipped off an outcrop in the Western Australian scrub turn out to agree on a number to five decimal places.
That number is a ratio of two strontium isotopes, and the agreement is the whole story. A team led by PhD student Matilda Boyce at the University of Western Australia has measured it in Australia’s oldest rocks and found that it matches the Moon.
What the outback rocks are
Samples came from the Manfred Complex in the Murchison region, part of the Narryer Terrane, and at 3.73 billion years old they are the oldest rocks on the Australian continent. They are anorthosites, which means they consist almost entirely of one mineral: plagioclase feldspar. On Earth that is a geological oddity. On the Moon it is standard issue, because the pale highlands visible with the naked eye are anorthosite crust.
How a feldspar crystal keeps time
So what turns a crystal into a clock? Rubidium-87 decays into strontium-87 with a half-life of roughly 50 billion years, so any mineral carrying rubidium slowly accumulates strontium-87 and its isotope ratio creeps upward. Plagioclase is the useful exception. It hoovers up strontium and admits almost no rubidium, so the ratio sealed in at crystallisation barely drifts afterwards.
Writing in Nature Communications, Boyce and her co-authors report an initial ratio of 0.700050 in Manfred plagioclase, the least radiogenic strontium ever measured precisely in any terrestrial sample.
The Moon rock that matches
During Apollo 16, astronauts John Young and Charles Duke collected sample 60025 in the Descartes highlands, a 1.8-kilogram lump of ferroan anorthosite almost within arm’s reach of the lander. The Open University’s Virtual Microscope describes it as a piece of the Moon’s original crust, formed when plagioclase floated to the top of a Moon-wide magma ocean.
Its plagioclase yields an initial strontium ratio of 0.699062, dated at 4.51 billion years old. Wind both bodies back to about 4.515 billion years ago, correcting for how much rubidium each actually carries, and Earth and Moon converge on 0.699061.
They started from the same pot.
Rubidium is the element that ran away
Rubidium is moderately volatile and strontium is not, which matters enormously in a collision hot enough to melt two planets. As Space.com has reported, the standard account holds that a Mars-size body slammed into the young Earth, flinging off the debris that became the Moon. Rubidium boils off. Strontium stays put.
The Moon came out of that event much poorer in rubidium than Earth, so its strontium clock has been running slow ever since. That asymmetry is not an awkward detail in the model; it is what the model predicts, and the measured ratios oblige.
Most of the crystal is useless
“We used fine-scale analytical methods to isolate the fresh areas of plagioclase feldspar crystals,” Boyce said in a statement released by her university. She was not being modest. Manfred rocks were cooked repeatedly by thermal events between roughly 3.7 and 1.8 billion years ago, and fluids shunted rubidium and strontium around inside the crystals, wrecking most of the record.
To find the survivors, the team first mapped a megacryst’s chemistry at ANSTO’s X-ray fluorescence microscopy beamline, marking where rubidium and strontium had migrated. They then measured actual strontium isotopes with a laser instrument, walking a grid of 110-micron spots across the same crystal. Fresh patches look optically dark. Altered, fractured patches look white, and they are rubidium-rich and radiogenic. The headline number rests on 35 analyses inside two small dark patches.
A second isotope system policed the result. Potassium-40 decays into calcium-40, and old continental crust, rich in potassium, becomes wildly radiogenic in calcium over time. A nearby gneiss of similar vintage sits nearly 2,000 parts per million above the reference standard. Manfred plagioclase came in at minus 45, indistinguishable from mantle, which rules out crustal contamination in the parent magma.
The continents turned up late
A depleted mantle leaves a strontium fingerprint, and it only shows up in the terrestrial record after about 3.5 billion years ago. Before about 4 billion years ago, the study found nothing at all. Continental crust got going roughly a billion years after the planet did.
That reading also demotes the previous record holder. Very unradiogenic strontium in Archean barites had long been used to anchor bulk Earth, and the new work reinterprets those values as seawater rather than mantle, which is why the authors propose starting the strontium depleted mantle curve at 3.8 billion years rather than 4.5.
One paper, one crystal
The giant impact hypothesis was already well supported before any of this. Oxygen, silicon, calcium and magnesium isotopes have pointed at a violent, high-energy merger for years, and the picture is fairly settled. Strontium was the holdout, with modelled starting values for Earth and Moon that stubbornly refused to line up.
This is one study, offering a reconciliation rather than a proof. Its foundation is two dark patches in one crystal from one intrusion in the Murchison, which is a remarkable place for the origin of the Moon to be hiding, and a thin place for it to stand. Somebody now has to go and find the second crystal.
