Earth’s tides pushed the Moon’s crust to its far side, Chang’e-6 rocks suggest

Look up at a full Moon and the dark patches you see are old lava plains, the familiar face people have watched for as long as there have been people. The hemisphere that never turns toward us looks like another world: pale, crowded with craters, almost bare of dark plains and wrapped in a much thicker crust. Rocks scooped from that hidden side by a Chinese lander now suggest the split was set by Earth, whose tidal heat pushed the young Moon’s floating crust toward the far side.

The finding goes to the root of one of the oldest puzzles in lunar science, and it rests on the first comparison ever made with real rocks from both hemispheres. Researchers at Nanjing University and the Chinese Academy of Sciences matched far-side fragments against samples from the near side, layer by layer, from the deep interior to the surface. Their answer is that the Moon is lopsided in its crust and in the last dregs of its ancient magma ocean, but not in the mantle underneath.

The far-side material comes from Chang’e-6, which landed in Apollo crater inside the vast South Pole-Aitken basin and returned 1,935.3 grams of soil and rock to Earth. Until then, every lunar sample in a laboratory, from the Apollo and Luna missions to China’s Chang’e-5, had come from the side we can see. The new study, published in National Science Review, reads the far side’s history from small rock fragments picked out of that soil.

How the far-side rocks were read

The team searched an allocated portion of Chang’e-6 soil and identified 47 rock fragments, known as clasts: 37 from the highland crust and 10 pieces of mare basalt, the dark volcanic rock that floods the near side. They mapped each clast’s texture in three dimensions, measured the chemistry of its minerals and dated the basalts from the decay of uranium into lead.

Each rock type samples a different layer of the Moon. Basalts form when the mantle partially melts, so they carry the mantle’s chemical fingerprint to the surface. Anorthosite, a pale rock made almost entirely of the mineral plagioclase, is the stuff of the original crust, which floated to the top of a global ocean of molten rock as the newborn Moon cooled. A third family, the Mg-suite, forms where melted mantle material pushes into that crust, so it records how the two layers interacted.

The yardstick was the near-side record: decades of measurements on Apollo rocks and on the basalts Chang’e-5 brought back from Oceanus Procellarum, the Ocean of Storms. The researchers also checked their numbers against orbital maps of both hemispheres, a way to test whether a handful of fragments from one crater looked typical of the far-side highlands.

Same mantle, different crust

The basalts delivered the first surprise. They are low-titanium basalts about 2.81 billion years old, and their trace elements and isotopes match the mantle sources of the low-titanium basalts collected by the Apollo astronauts. Their source has one of the lowest uranium-to-lead ratios measured in any returned lunar basalt, yet it sits on the same chemical path as its near-side cousins.

The Mg-suite rocks told the same story from another angle. Their magnesium number, a ratio of magnesium to iron that tracks the mantle material they came from, peaks at about 79 on both sides of the Moon, with identical medians of 78.0. If the deep interior had split into two different halves, these numbers should differ. They do not.

The crust is another matter. Pyroxene crystals in the far-side anorthosites peak at a magnesium number of about 67, against about 55 for the near-side highlands seen from orbit, and the gap held up statistically against the Apollo anorthosites, with medians of 66.4 versus 63.0. The far-side Mg-suite rocks are also strikingly poor in KREEP, the residue rich in potassium, rare-earth elements and phosphorus that is concentrated under the near side. Typical Apollo Mg-suite rocks are thought to contain 10 to 25 percent of that material by weight.

How Earth could have built the far side

The model starts with a Moon covered by a magma ocean that first cooled evenly, laying down mantle crystals and a uniform early crust. Roughly 20 million years after the Moon formed, it became tidally locked, keeping one face permanently toward a much closer Earth. Tides then heated the near side more than the far side, and that temperature gap set the remaining magma flowing sideways beneath the crust, a circulation the authors estimate may have lasted about 200 million years.

Because the young crust was still soft, the flow could drag it. Slabs of early, magnesium-rich anorthosite were carried toward the far side and piled up there, which would explain both its higher magnesium numbers and its thicker crust. The last and most evolved melt, rich in titanium and KREEP, drained the other way and pooled beneath the near side. That residue later fed the great outpourings of lava that cover about 31 percent of the hemisphere we see and about 1 percent of the far side.

The heat gap seems to have lasted. Measurements of how the Moon flexes under Earth’s monthly tides, together with basalt crystallization temperatures, indicate that the far-side mantle runs about 100 to 200 kelvin cooler than the near side. The model even offers a new reading of the South Pole-Aitken basin’s oval outline: instead of an oblique impact, the drifting crust may have stretched a basin that was once round.

What the rocks do not settle

The authors present their scheme as “a plausible and self-consistent explanation”, not as proof. All the far-side evidence comes from a single landing site, and the team acknowledges that the clasts “may not be representative of the entire farside” and that sampling bias cannot be entirely ruled out. The higher magnesium numbers in the far-side crust fit the model, but the paper notes that this match “does not provide a unique explanation” for the pattern.

The KREEP signal is open too. Low KREEP in the far-side Mg-suite could mean the layer really is thin on that side, that these rocks formed by a route that never needed KREEP, or that the layer is simply patchy across the Moon. The authors write that the current Chang’e-6 dataset cannot decide between these options and that the lopsided layering in their model “requires verification with additional farside samples.”

Rival explanations stay on the table: a giant impact that redistributed material, uneven growth of the crust, and extra heat from a smaller companion moon that merged with ours. What the study adds is a constraint every one of them now has to meet, a mantle that looks chemically alike under both faces.

Common questions about the Moon’s far side

Why does the far side of the Moon look so different from the near side?

The near side is covered by dark, smooth basalt plains, while the far side is almost all bright, cratered highland with a thicker crust. The Chang’e-6 study proposes that tidal heat from Earth drove sideways flow in the young Moon’s magma ocean, piling crust onto the far side and pushing lava-forming material toward the near side.

Is the far side of the Moon always dark?

No. The far side receives as much sunlight as the near side; it is only hidden from Earth because the Moon is tidally locked and always shows us the same face. At new Moon, the far side is fully lit.

What did Chang’e-6 bring back from the Moon?

China’s Chang’e-6 returned 1,935.3 grams of soil and rock from Apollo crater, inside the South Pole-Aitken basin, in June 2024. They are the first samples ever collected from the lunar far side.

Why does the Moon always show the same face to Earth?

Earth’s gravity raised bulges in the young Moon, and tidal friction slowed its spin until one rotation matched one orbit. The Nanjing team’s model places this at roughly 20 million years after the Moon formed, when Earth was much closer and tidal heating much stronger.

The paper appeared online in National Science Review on 13 July 2026, and Science China Press released a summary on 14 October. The authors suggest the same sideways convection could shape any world that keeps one face toward its host, and may help explain uneven ice-shell thickness on Jupiter’s moon Europa and Saturn’s moon Titan. On the Moon itself, the test needs more far-side rock, and China’s next lunar mission, Chang’e-7, is bound for the south pole rather than the far side and missed its 2026 launch window.

Reference: Gu et al., “Lunar farside-nearside asymmetry formation during magma ocean solidification,” National Science Review, 2026. DOI: 10.1093/nsr/nwag424

Tags: astronomy, moon, planetary science, Chang'e-6, far side of the Moon, Nanjing University