The Moon is too big. Relative to Earth, it is the largest satellite of any rocky planet in the solar system — roughly 1/80th of Earth's mass, compared to a ratio of about 1/4,000 for Mars and Phobos. A Moon this large, this close, with these chemical properties, cannot be explained by any of the formation scenarios that work for other planetary satellites. It was not captured from elsewhere (the orbital mechanics don't work). It did not form alongside Earth from the same cloud (the chemistry is wrong). Something unusual happened early in solar system history, and the best-supported explanation involves a violent collision between the proto-Earth and another protoplanet during the era of planet formation, about 4.5 billion years ago.
The giant impact hypothesis was proposed independently in 1975 by William Hartmann and Donald Davis and by A.G.W. Cameron and William Ward. It posits that a Mars-sized body, conventionally named Theia, struck the young Earth at an oblique angle. The energy of the collision partially vaporized both bodies. The iron cores of Theia and proto-Earth merged, while a disk of silicate vapor, melt, and debris was ejected into orbit around the reformed Earth. That disk, within tens of millions of years, accreted into the Moon.
What the chemistry says
The hypothesis explains several key features of the Moon that other formation scenarios cannot. The Moon is depleted in iron relative to Earth, because its material comes mostly from the outer silicate layers of both bodies, while the iron settled into the merged core. The Moon is depleted in volatile elements — water, sodium, potassium, sulfur — relative to Earth, because the high-energy collision and the subsequent hot disk environment boiled off volatiles that escaped into space. The Moon's density is lower than Earth's for the same reason: it is mostly silicate, with very little iron.
The most critical test involves oxygen isotopes. Isotope ratios in planetary bodies are characteristic of where they formed in the solar system. If Theia formed far from Earth and struck at high velocity, it should have a different oxygen isotope ratio, and the Moon — which should be largely made of Theia's material — should have a different ratio from Earth. But it does not. The oxygen isotope ratios of Earth and Moon rocks are essentially identical, to a precision that has been measured in multiple laboratories. This is the central problem with the standard giant impact model: it predicts a Moon compositionally similar to Theia, but the Moon looks like Earth.
The synestia and alternative models
Several modifications to the original hypothesis have been proposed to resolve this. One is that Theia happened to form in the same part of the solar system as Earth and therefore had the same isotopic composition — a coincidence that is plausible but unsatisfying. Another involves a higher-energy, more head-on impact that vaporized and mixed both bodies so thoroughly that the resulting debris cloud was isotopically homogeneous. A more radical alternative, proposed by Sarah Stewart and Simon Lock in 2018, involves a "synestia": a rapidly spinning, mostly vaporized, donut-shaped mass of Earth-derived material in which the Moon forms from a condensing vapor cloud rather than from a disk of solid debris. The synestia model naturally produces a Moon with Earth's isotopic signature because it forms entirely from Earth's own vaporized mantle material, with Theia's contribution thoroughly mixed in.
Apollo samples continue to generate new insights. High-precision measurements of stable titanium, chromium, and tungsten isotopes have progressively refined estimates of how much Theia's material is present in the Moon and how thoroughly the two bodies mixed. The Chinese Chang'e-5 samples, returned in December 2020 from a younger volcanic region than any Apollo landing site, have added data points on lunar volcanic history and water content that are still being analyzed. One finding from those samples: lunar mantle water content appears higher than many post-Apollo models predicted, suggesting either that volatile delivery to the early Moon was more efficient than thought or that the Moon started with more water than the giant impact would have left behind.
The Moon's origin is not a closed problem. It is one of those questions where the broad answer — a large impact — is well established, but the details resist resolution by any single model. The next major data return, from Artemis sample-return missions targeting the lunar south pole, may add a different chapter: what the permanently shadowed craters contain, how that ice got there, and what it says about the delivery of volatiles to the inner solar system in the aftermath of the Moon's violent birth.
