Earth’s earliest continents, formed in the Archean eon, roughly 4–2.5 billion years ago, weren’t just pristine melts straight from the mantle. They incorporated a lot of “sun- baked ocean leftovers”—recycled oceanic crust and sediments that had been altered at the surface by seawater and the ancient atmosphere.

Researchers studied ~2.7–2.5 billion-year-old granitic rocks (part of the tonalite-trondhjemite-granodiorite or TTG suite, the main component of early continental crust) from China’s North China Craton.

They looked at:

Sulfur isotopes (especially mass-independent fractionation, Δ³³S): In the oxygen-poor Archean atmosphere, UV light from the Sun created distinctive sulfur anomalies in surface materials. These signatures don’t form deep in the mantle and survived in the rocks.

Silicon isotopes: Seawater alteration of basalt or sedimentary processes make silicon “heavier” (higher proportion of heavier isotopes) compared to mantle-derived material.

The combination of these signatures in the granites points strongly to supracrustal sources—rocks that had been at or near the surface, interacted with the ocean and atmosphere, then got recycled into magma sources that melted to form new continental crust. Pure mantle melts wouldn’t show this double fingerprint.

This pattern holds across global Archean records after ~3.8 billion years ago, suggesting surface recycling was a dominant process in building the first stable continents.

It’s a nice reminder that Earth’s rock cycle —erosion, sedimentation, alteration, subduction/melting, and uplift—has been running for billions of years, long predating human environmentalism. The continents we stand on today carry atoms that once sat on ancient seafloors under a hazy, anoxic sky.

_____________________________________________________________________________________

Coupled sulfur-silicon isotopes reveal supracrustal origin of Archean continents

This study uses coupled whole-rock quadruple sulfur (δ³⁴S, Δ³³S, Δ³⁶S) and silicon (δ³⁰Si) isotopes from Neoarchean (~2.7–2.5 Ga) granitoids in the Luxi area (North China Craton) to argue that the dominant source for preserved Archean continental crust was recycled supracrustal mafic material (seawater-altered oceanic basalts/sediments) rather than pristine mantle-derived mafic cumulates or lower-crustal gabbros.

Archean continental crust (mainly the tonalite-trondhjemite-granodiorite or TTG suite) formed primarily by partial melting of supracrustal mafic sources—oceanic basalts and related materials that had been altered at the surface by seawater and the anoxic Archean atmosphere—rather than pristine, unaltered mafic cumulates from the mantle or lower crust.

Single-isotope systems have ambiguities:

Zircon δ¹⁸O, often mantle-like in these rocks, can be reset or buffered and doesn’t rule out minor supracrustal input.

δ³⁰Si alone: Heavy values (elevated δ³⁰Si) suggest seawater silicification or sedimentary silica addition but could involve other processes.

Quadruple S (MIF-S): Non-zero Δ³³S is a robust Archean atmospheric UV-photochemistry fingerprint (SO₂ reactions in anoxic air). It survives high-T processing but can be subtle.

Coupling them is key. Sulfur (volatile trace element) and silicon (major element) behave differently during metamorphism/melting due to contrasting rock-fluid partitioning (higher Damköhler number for Si → rock-buffered over short distances; S equilibrates over longer paths). Magmatic/metamorphic processes are unlikely to fractionate them congruently, so covariation strongly supports inheritance from a shared supracrustal precursor.

Data from Luxi:

• Granitoids: Small but resolvable non-zero Δ³³S (down to −0.24‰, avg. ~−0.06‰) outside typical mantle range; δ³⁴S mostly near 0; Δ³⁶S negative.
• δ³⁰Si: Elevated/heavy (−0.09‰ to −0.05‰ in ~2.7–2.6 Ga; more variable later).
• Contrasts with previously reported mantle-like zircon δ¹⁸O in the same region.
• Amphibolites (mafic) also show some MIF-S.

These TTGs are mostly medium- to high-pressure (deeper melting), calc-alkaline, sodic, with TTG-like REE patterns.

Global compilation and implications

After ~3.8 Ga, most Archean granitoids worldwide show this paired signature: enriched δ³⁰Si + non-zero Δ³³S. This implies supracrustal recycling was the dominant pathway for building stable felsic crust—not just a minor contributor. Pure juvenile mantle melts or unaltered cumulates wouldn’t carry these surface fingerprints.

Implications

• Supports models involving hydrothermal alteration of oceanic crust (e.g., in subduction-like or plateau settings) followed by melting to produce TTGs.
• Indicates early crustal recycling (proto-plate tectonics or vigorous vertical tectonics/volcanism) was active by the late Archean, coupling atmosphere-ocean-interior processes.
• Ties into early habitability: Surface-interior exchange helped regulate volatiles, nutrients, and conditions for life.

The study doesn’t claim zero juvenile mantle input, but the dominant pathway for preserved Archean felsic crust involved supracrustal precursors. It strengthens the “recycling Earth” narrative you highlighted earlier.

This strengthens the “Earth as recycler” view from the start of the rock cycle we recognize today. The continents aren’t pristine mantle additions but reworked “sun-baked ocean leftovers,” as your original headline put it. It refines geodynamic models without fully settling the plate tectonics timing debate.

Published: Nature Communications

DOI: 10.1038/s41467-026-72701-4

Authors: Kun Shang, 
Jian Zhang, 
Zaicong Wang, 
Ian Cawood, 
Yawen Cui, 
Ming Li, 
Ruihong Chang, 
Yanan Shen & 
Guochun Zhao

Abstract

The genesis of Archean continental crust through partial melting of hydrous mafic protolith is widely acknowledged, yet the origin of the mafic protolith remains highly contentious. Silicon isotopes and quadruple sulfur isotopes serve as particularly powerful tools in this regard, as they directly trace the source nature of the felsic continent. Here, we integrate whole-rock silicon and sulfur isotopic data of Neoarchean granitoids from the North China Craton to constrain the origin and pathway of their protoliths. These granitoids exhibit non-zero Δ³³S (0.01‰ to − 0.24‰) and elevated δ³⁰Si ( − 0.09‰ to −0.05‰), requiring a supracrustal source and contrasting with the previously reported mantle-like zircon δ¹⁸O. Global compilation shows that granitoids formed after 3800 Ma uniformly contain enriched δ³⁰Si and non-zero Δ³³S, implying that most, if not all Archean continental crusts were derived from partial melting of supracrustal sources rather than unaltered mafic cumulates.