The Giant-Impact Hypothesis

Why the giant-impact hypothesis matters

The Moon is not a captured asteroid or a leftover chip of a molten Earth — it is the frozen record of a planetary catastrophe. Before the 1970s, three older models competed: fission (the fast-spinning Earth flung off a blob that became the Moon), capture (the Moon formed elsewhere and was gravitationally trapped), and co-accretion (Earth and Moon grew side by side from the same disk). Each failed a hard test. Fission demands an implausibly large initial spin — far more angular momentum than the present Earth-Moon system has, with no way to shed the excess. Capture requires an implausibly precise, energy-shedding encounter. Co-accretion cannot explain why the Moon is so iron-poor compared with Earth. The giant impact succeeds where all three fail, which is why it has been the consensus model since the 1984 Kona conference on lunar origin.

• It solves the density paradox. The Moon's mean density of 3.34 g/cm³ is close to that of Earth's mantle, not the whole Earth — exactly what you expect if the Moon is made of ejected mantle rock and almost no iron.
• It supplies the angular momentum. An off-center collision spins up the system to the ~3.5×10³⁴ kg·m²/s we measure today, a value nothing else easily reproduces.
• It sets the geochemical clock. Hafnium-tungsten dating of lunar samples ties the impact to roughly 30-100 million years after the first solids condensed, anchoring the timeline of terrestrial-planet formation.
• It shaped habitability. The Moon stabilizes Earth's axial tilt and drives ocean tides — both plausibly relevant to climate and the origin of life.

How it works, step by step

The canonical (Cameron-Benz-Canup) scenario proceeds like this:

1. Two differentiated bodies. By the time of impact, both the proto-Earth (~0.9 M⊕) and Theia (~0.1 M⊕) had already melted and separated into iron cores and silicate mantles. This differentiation is essential — it is what lets a grazing hit skim off iron-poor mantle.
2. An oblique, low-speed collision. Theia strikes at roughly 9-10 km/s (close to the Earth-Theia mutual escape velocity of ~9-11 km/s, so barely faster than free-fall) at an angle near 45°. A head-on hit would merge the bodies with too little debris; a grazing hit maximizes the orbiting disk.
3. Core merger. Theia's dense iron core plunges through the proto-Earth's mantle and sinks, merging with Earth's core within hours. This is why the Moon inherits so little iron.
4. Disk formation. Shocked, vaporized and molten silicate from the outer layers of both bodies is flung outward. A fraction with enough angular momentum settles into a partly-vapor, partly-liquid protolunar disk extending past the Roche limit (~2.9 Earth radii for lunar-density material).
5. Rapid lunar accretion. Material beyond the Roche limit coagulates. Simulations show a single dominant moon assembling in as little as decades to a few thousand years, initially just ~3-5 Earth radii away — roughly 12-20 times closer than today's ~60 Earth radii.
6. Magma ocean and differentiation. The newborn Moon is molten. As its global magma ocean cools over tens of millions of years, low-density plagioclase floats to form the bright anorthositic highland crust while dense minerals sink — the classic lunar magma-ocean model.
7. Tidal recession. Tidal friction then transfers Earth's spin into the Moon's orbit, driving lunar recession at about 3.8 cm/year and lengthening Earth's day from an initial ~4-5 hours toward 24 hours.

Key numbers: Earth versus Moon

Property	Earth	Moon	What it tells us
Mass	5.97×10²⁴ kg	7.35×10²² kg (0.0123 M⊕)	Moon is ~1/81 of Earth
Mean radius	6371 km	1737 km	Moon ~27% of Earth's radius
Bulk density	5.51 g/cm³	3.34 g/cm³	Moon ≈ Earth's mantle, iron-poor
Core (fraction of mass)	~32%	~1-2%	Grazing impact stripped iron
Δ¹⁷O (oxygen isotopes)	0 (reference)	within ~1-6 ppm of Earth	Same reservoir — the isotopic puzzle
Bulk volatiles (e.g. K, Na, H₂O)	Enriched	Strongly depleted	Baked out in the hot debris disk

A key equation: the angular-momentum budget

The total angular momentum of the Earth-Moon system is the sum of Earth's spin and the Moon's orbit:

L = I_⊕ ω_⊕ + m_☾ √(G M_⊕ a) ≈ 3.5 × 10³⁴ kg·m²/s

where:

• I_⊕ = Earth's moment of inertia ≈ 8.0×10³⁷ kg·m² (using I = 0.331 M_⊕ R_⊕²);
• ω_⊕ = Earth's spin angular velocity (7.29×10⁻⁵ rad/s today);
• m_☾ = Moon's mass = 7.35×10²² kg;
• G = gravitational constant = 6.674×10⁻¹¹ m³ kg⁻¹ s⁻²;
• M_⊕ = Earth's mass = 5.97×10²⁴ kg;
• a = Earth-Moon distance = 3.844×10⁸ m (semi-major axis).

Roughly 80% of that total lives in the Moon's orbital term. Angular momentum is conserved, so tidal friction can only shuffle it between Earth's spin and the Moon's orbit — it cannot create or destroy it. Run the clock backwards and Earth spun with a ~4-5 hour day while the Moon hugged close. The giant impact must have delivered this entire budget in one blow, and reproducing the value ~3.5×10³⁴ kg·m²/s is the single tightest constraint any Moon-formation model must satisfy. The Roche limit that sets where the disk can coalesce is

d = 2.44 R_⊕ (ρ_⊕ / ρ_☾)^1/3 ≈ 2.9 R_⊕ ≈ 18,400 km,

where ρ_⊕ and ρ_☾ are the bulk densities of Earth and Moon and the 2.44 coefficient is the fluid-satellite value (a rigid body survives closer in). Debris inside this radius is tidally shredded; only material flung beyond it can accrete into a moon.

The catch: the isotopic crisis and the synestia

The classic model has one stubborn flaw. Bodies that formed at different distances from the Sun carry different oxygen-isotope fingerprints — Mars, Vesta and the various chondrite classes all differ measurably in their Δ¹⁷O values. Simulations of the canonical impact put roughly 60-80% of the Moon's mass into Theia's material. Yet lunar and terrestrial rocks match to within a few parts per million in oxygen (and also in titanium-50, chromium-54, tungsten-182 and more). If most of the Moon is Theia, why does it look exactly like Earth's mantle? This is the isotopic crisis.

Two fixes dominate. The first is higher-energy, faster-spinning impacts (Ćuk & Stewart 2012) that eject a disk dominated by Earth material, but which then require the system to shed excess angular momentum through a later solar-tidal resonance (the evection resonance). The second is the synestia (Lock & Stewart 2017): if the collision vaporizes enough rock and spins fast enough, Earth and impactor merge into a single, connected, donut-shaped cloud of silicate vapor with no distinct surface, thousands of kilometers across. Inside this synestia, Earth-derived and Theia-derived vapor mix thoroughly before the Moon condenses out from within over roughly a decade — homogenizing the isotopes automatically. The synestia elegantly turns the isotopic crisis from a bug into a natural prediction.

The late veneer: a final sprinkle of gold

The story does not end when the Moon forms. Earth's mantle contains more highly siderophile (iron-loving) elements — gold, platinum, iridium, osmium — than it should, because during core formation those elements should have followed the iron down to the core. The resolution is late accretion, sometimes called the late veneer: after core formation was complete, both Earth and the Moon kept sweeping up leftover planetesimals, delivering ~0.5% of an Earth mass of chondritic material that stranded these metals in the mantle. Crucially, Earth's far larger gravitational cross-section means it accreted roughly 20 times more late material per unit mass than the Moon, which explains why the two bodies differ in their siderophile abundances despite forming together. The late heavy bombardment that scarred the lunar highlands is a still-later, distinct episode.

Common misconceptions

• "Theia hit Earth head-on." No — the impact was grazing and oblique (~45°). A head-on hit would merge the two bodies and leave too little orbiting debris to build a Moon.
• "The Moon is a chunk of Earth's crust." Not quite. It is mantle-derived rock from both Theia and Earth, processed through vapor and magma, not a cleanly detached piece of pre-existing crust.
• "It took millions of years for the Moon to form after the impact." The collision and disk-to-Moon accretion were fast — decades to a few thousand years. What took tens of millions of years was the cooling and crystallization of the lunar magma ocean.
• "The Moon formed at its current distance." It formed just ~3-5 Earth radii away and has been receding ever since; today it sits at ~60 Earth radii.
• "The impact was faster than a comet or asteroid strike." On the contrary, ~9-10 km/s is slow for a Solar-System collision — barely above mutual escape velocity. That gentleness is exactly why so much material stayed bound in orbit rather than escaping.
• "We've proven Theia existed." Theia is a well-motivated inference, not a directly observed object. Some researchers even argue two dense blobs deep in Earth's mantle (the LLSVPs) may be buried remnants of Theia — an intriguing but unconfirmed idea.

A short history

The modern hypothesis emerged independently twice in the mid-1970s: William Hartmann and Donald Davis (1975) argued that planetesimal accretion should produce Mars-sized bodies capable of catastrophic impacts, while Alastair Cameron and William Ward (1976) showed that a Mars-sized impactor could supply the Earth-Moon angular momentum. The idea languished until the 1984 Kona conference on the origin of the Moon crowned it the leading model. The first smoothed-particle-hydrodynamics (SPH) simulations by Benz, Slattery and Cameron in the late 1980s made it quantitative, and Robin Canup's work in the 2000s produced disks that reliably yield a lunar-mass moon. The isotopic-crisis papers (Ćuk & Stewart 2012; Canup 2012) and the synestia model (Lock & Stewart 2017) are the current frontier — the giant impact is not one fixed story but a maturing family of collision scenarios.