A low belt of dark, weathered rock beside Hudson Bay contains something almost absent from Earth: a physical sequence that may have survived from the planet’s first eon. The Nuvvuagittuq Greenstone Belt, near Inukjuak in northern Quebec’s Nunavik region, is made largely of metamorphosed volcanic rocks rich in magnesium and iron.
A 2025 study in Science dated mafic intrusions inside the belt to the Hadean, about 4.16 billion years ago. Those intrusions cut through an older volcanic unit. The visible sequence therefore turns the isotope result into a minimum age: the surrounding rocks had to exist before the magma entered them.
This is one study, not settled consensus. It provides the strongest evidence yet that Nuvvuagittuq contains Hadean rocks, but it does not give every stone at the site one exact age. Some geochemists remain cautious about the uncertainties and would like confirmation from additional radiometric systems.
The oldest rock is not the oldest terrestrial material
The superlative needs a definition. Zircon crystals from the Jack Hills of Western Australia formed far earlier, including one dated to about 4.404 billion years. A landmark analysis of that zircon found chemistry consistent with early crust and liquid water.
A crystal and a rock body answer different questions. The crystal carries a durable chemical record of the magma in which it grew. A mapped outcrop can additionally preserve contacts, relative order, deformation and evidence that one event happened before another.
Those zircons are the oldest known pieces of terrestrial material, but not intact rocks from their time. Their original host rocks were destroyed. The resilient crystals survived erosion and were later incorporated into sedimentary rocks hundreds of millions of years younger.
Nuvvuagittuq offers a different archive. Its ancient material remains in a coherent, mappable body, allowing geologists to examine which units touch, intrude or predate one another. Before the new work, the roughly 4.03-billion-year-old Acasta Gneiss in Canada’s Northwest Territories was generally treated as the oldest undisputed intact rock.
Here, “intact” does not mean pristine. The belt has endured more than four billion years of heating, pressure, folding, deformation and fluid movement. It means enough of the rock body and its field relationships remain together to reconstruct a sequence of events rather than relying on a detached mineral grain.
A 2008 date opened a long argument
The record claim began nearly two decades ago. SpaceDaily’s 2008 report from Nuvvuagittuq described whole-rock dates ranging from 3.8 to 4.28 billion years and identified a unit then called faux amphibolite as ancient volcanic material.
The underlying Science paper used samarium-146 and its daughter neodymium-142. Samarium-146 was present when the Solar System was young but decayed away early in Earth’s history. Variations in neodymium-142 can therefore retain information about chemical reservoirs established during the Hadean.
The difficulty was deciding what the signal dated. It could record crystallisation of the rocks themselves, but it could also be inherited from an older mantle or crustal source that was melted to make younger rocks. A chemical memory from Hadean material is not automatically a Hadean formation age for the sample holding it.
Other evidence pointed younger. A 2012 geochronology study combined samarium-neodymium systems with uranium-lead zircon ages and described a history extending over more than 1.5 billion years. Zircon-bearing intrusions established that at least some events occurred around 3.8 billion years ago or later. The belt plainly did not form or change only once.
Zircon-poor basalt required another kind of clock
Zircon is valuable because uranium enters its crystal structure when it forms while lead is largely excluded. Two uranium isotopes then decay into two lead isotopes at known rates, providing paired clocks that can survive many geological disturbances.
Nuvvuagittuq’s dominant rocks began as mafic or ultramafic volcanic material. Such magnesium- and iron-rich compositions generally contain too little zirconium to crystallise abundant zircons. The mineral that provides the standard test for the Acasta Gneiss and Jack Hills is scarce in the unit whose age matters most.
That leaves whole-rock isotope methods. Instead of isolating one crystal, researchers analyse related samples whose parent-to-daughter element ratios differ because magma separated into distinct compositions. If the samples formed together and remained a meaningful system, their isotope ratios define an isochron whose slope corresponds to elapsed time.
The method can be powerful, but ancient rocks have been reheated, metamorphosed and exposed to fluids. Later events can move elements, partially reset a clock or combine material with different histories. The geological relationship among samples is therefore as important as the numbers returned by a mass spectrometer.
The team dated rock that cuts older rock
The 2025 team changed the target. Samples collected in 2017 came from metagabbroic intrusions that visibly cut the belt’s dominant Ujaraaluk unit, a name translated by the researchers as “big old solid rock.” The once-molten material forced its way through pre-existing volcanic rock and crystallised.
The logic is basic relative geology: an intrusion must be younger than the rock it penetrates. Dating the intrusion therefore places a floor beneath the host’s age. If the crosscutting metagabbro crystallised during the Hadean, the Ujaraaluk sequence was already present during the Hadean.
Carleton University geochemist Hanika Rizo, a co-author, gives a field account of the sampling strategy. The belt lies roughly 30 kilometres from Inukjuak, north of the 55th parallel. Its outcrops preserve contacts that link laboratory measurements to a physical ordering visible on the ground.
This distinction is why “at least 4.16 billion years” is defensible for the older host rocks. The paper did not obtain one exact crystallisation age for the entire outcrop. It dated an intrusive event, then used crosscutting relationships to establish that the enclosing rocks are older.
Two samarium-neodymium clocks reached the Hadean
The long-lived samarium-147 to neodymium-143 system yielded an isochron age of 4,157 ± 174 million years. The extinct samarium-146 to neodymium-142 system yielded 4,196 million years, with an interval extending 81 million years younger and 53 million years older.
The two central values are not identical, and the uncertainty on the long-lived system is broad. Their ranges overlap, however, and both are consistent with Hadean crystallisation. The short-lived system’s reported interval remains older than the roughly 4.03-billion-year boundary used in the paper.
These are complementary chronometers rather than completely unrelated clocks. Both depend on the behaviour of samarium and neodymium in the same suite of fractionated rocks. Their parent isotopes have radically different half-lives, however, so later heating and chemical disturbance do not affect their time information in the same way.
Agreement therefore makes a common inherited source or a simple reset explanation harder to sustain. It does not make alternative histories mathematically impossible. Independent coverage in Chemical & Engineering News noted that some specialists remain unconvinced, citing possible geophysical effects and the large uncertainty of the long-lived date.
“Only known” is a present boundary, not a timeless fact
If the interpretation holds, Nuvvuagittuq is the only known coherent rock fragment preserved from Earth’s first roughly 500 million years. That wording describes the current geological archive. It does not show that this was the only crust then existing, nor that another Hadean outcrop cannot be recognised later.
The distinction also separates preserved rock from preserved chemistry. Ancient isotope anomalies can survive in younger lavas, revealing reservoirs established near Earth’s formation. Detrital zircons preserve grains from vanished crust. Meteorites and lunar samples record Solar System history older than most Earth rocks. None supplies the same package of neighbouring terrestrial units and readable field contacts.
SpaceDaily’s report on using ancient zircons to reconstruct Hadean crust shows what researchers can infer when the parent rocks are gone. Trace elements inside the crystals act as proxies for magma composition, but the reconstruction depends on relationships learned from younger rocks.
Nuvvuagittuq can potentially test such reconstructions against surviving rock. Its mafic volcanic units, intrusions and chemically altered layers may constrain how early crust differentiated, how mantle reservoirs separated and how water interacted with rock. “Intact” supplies context, not immunity from interpretation.
A Hadean rock is not an untouched Hadean landscape
Earth formed about 4.54 to 4.57 billion years ago. The Hadean ended around 4.03 billion years ago under the boundary used by the study. During that interval the planet assembled, differentiated internally, experienced the Moon-forming collision and continued to receive large impacts while its surface and atmosphere evolved.
The name can encourage an image of uninterrupted inferno, but zircon evidence indicates that crust and liquid water appeared early. SpaceDaily’s analysis of water partitioning during magma-ocean cooling describes another part of that transition: the young planet could distribute water between its solidifying mantle, atmosphere and eventual surface reservoirs.
Nuvvuagittuq has not preserved a frozen scene from that surface. The original basaltic and associated rocks were metamorphosed, folded and chemically modified. Researchers must decide which features are primary, which came from later fluids and which were reset during younger heating.
The age alone also does not demonstrate that life existed there. Some units formed through interaction with water and may retain clues to early environments, but a Hadean date is not a biosignature. The paper establishes a geological clock and ordering, not a fossil, metabolic product or inhabited ecosystem.
The oldest archive is also a protected place
The belt is on Inuit-owned land near Inukjuak, not an unlimited global sample cabinet. Its scientific fame has already produced physical damage. Local landholders restricted access after large pieces were removed and specimens attributed to the site appeared for sale.
Associated Press reporting describes the Pituvik Landholding Corporation’s effort to protect the formation while exploring a provincial park and controlled research access. The community’s permission and priorities are part of any future field programme.
This is not separate from scientific quality. An isotope result is only as reliable as its sample context. Careful mapping, documented collection and restraint preserve the field relationships needed to test an extraordinary age. Removing a block without context can turn an irreplaceable part of the archive into little more than a curiosity.
Further work may narrow the dates or develop a model that changes their interpretation. For now, two samarium-neodymium systems and a crosscutting intrusion converge on the same consequential point: at least part of the rock exposed beside Hudson Bay was already there during the Hadean. The fragment is battered rather than pristine, but it carries a geological sequence from a chapter almost entirely erased elsewhere.