A wedding ring weighs a few grams. The gold in it has been mined, refined, alloyed, cast, polished, and fitted. Almost every step of that process happened on Earth in the last few hundred years. But the gold itself, every atom of it, was made somewhere else, in conditions that no longer exist anywhere in the solar system, and it was made before there was an Earth to mine it from.
This is not just a poetic claim. It is what the working models of nucleosynthesis say, and what direct observations over the past decade have started to confirm.
What stars can and cannot do
Hydrogen and helium come almost entirely from the first few minutes after the Big Bang, with a small amount of lithium. Everything else has to be assembled later. Most of what makes up the periodic table up to iron is built inside stars, by fusion, during their lives and in the supernovae that end them. This is the familiar “we are made of star stuff” point, and it is true for carbon, oxygen, calcium, iron, and most of the other elements that make up a human body.
Past iron, fusion in stars becomes a losing proposition. Assembling heavier nuclei costs energy rather than releasing it. Roughly half of the elements heavier than iron are made via the rapid neutron-capture process, the r-process, which requires something more extreme: neutron-rich material being ejected violently at high speed and high temperature, so that nuclei can absorb neutrons faster than they decay. Gold sits squarely in this category, alongside platinum, uranium, europium, and a long list of less familiar elements.
The question of where the r-process actually happens has been open in astrophysics for more than fifty years.
What 2017 showed
The clearest answer so far came on 17 August 2017, when LIGO and Virgo detected gravitational waves from the merger of two neutron stars in the galaxy NGC 4993, roughly 130 million light-years away. The event, designated GW170817, was followed within hours by a flash of gamma rays and within days by a fading glow visible across the optical and infrared, the signature of a “kilonova” powered by the radioactive decay of freshly forged heavy nuclei.
The spectroscopic analysis came in a Nature paper led by Elena Pian, which found spectral evidence consistent with r-process nucleosynthesis in the kilonova ejecta. A companion Nature paper by Daniel Kasen and colleagues modelled how much heavy-element material had been ejected. Total estimates landed near 16,000 Earth masses of heavy elements from the single event, with roughly ten Earth masses of gold and platinum in that figure.
GW170817 is, so far, the only neutron star merger to have been followed across the electromagnetic spectrum. It moved the production of r-process elements in such mergers from theoretical prediction to confirmed observation.
What 2025 added
A separate question is whether neutron star mergers alone can account for all the gold in the galaxy, or whether other events also contribute. Mergers are rare, and there is more r-process material in some old, metal-poor stars than the timing of merger events comfortably explains.
An April 2025 paper in The Astrophysical Journal Letters, led by Anirudh Patel of Columbia University, argued that magnetar giant flares are also a source. Magnetars are young, highly magnetised neutron stars; the giant flares they occasionally release are among the brightest transients ever recorded. The Patel paper revisited a delayed gamma-ray signal that followed the December 2004 flare from the magnetar SGR 1806-20, and found that the signal’s shape was consistent with the radioactive decay of freshly made r-process elements. The team estimated that such flares could account for up to ten per cent of the heavy elements in the Milky Way.
The result is recent. The contribution from magnetar flares is still being constrained, and the question of how the r-process budget is split between mergers, magnetar flares, and earlier-proposed sources is not settled. What is settled is that the gold in the solar system came from events of this general kind, and not from anything that has ever happened on Earth.
The arithmetic of age
Earth is about 4.54 billion years old, a figure derived from radiometric dating of meteorites and the oldest terrestrial mineral grains. The solar system itself, dated from calcium-aluminium inclusions in primitive meteorites, is slightly older, at about 4.567 billion years. The solar system formed when a giant molecular cloud collapsed under its own gravity, and that cloud already contained the heavy elements made in earlier events.
Every atom of gold now bound up in an Earth-bound wedding ring ultimately came from that pre-solar inheritance, material already enriched before Earth formed. The atoms are at least as old as the solar system, and almost certainly older, since the events that produce r-process elements happen across galactic history and the gold in any particular ring is a mixture of contributions from many of them.
This means the gold is older than the Pacific, older than the Atlantic, older than any continent now on Earth, older than the first cells that produced free oxygen, older than the dinosaurs, older than the first multicellular organisms, older than the first life of any kind we have evidence of. The ocean basins, the continents, the biosphere, the whole geological and biological history of the planet sits inside the age of the metal.
Holding it
The strange thing about this is how ordinary the object is. A ring is small enough to lose down a drain. Its provenance papers, if any, go back a few decades. Its actual provenance goes back to a stellar event nobody saw, in a galaxy that may or may not still exist in recognisable form.
What we keep returning to, in pieces like this, is that the very largest events in physics are not far away. The atoms produced in the most violent processes known to astronomy are, routinely, on people’s hands. The pleasing weight of the metal is, by its history, a piece of one of those events sitting quietly on a finger.
