Pop III

Population III

The First
Stars

For a hundred million years after the Big Bang the universe was dark. Then, inside knots of hydrogen and helium holding no trace of any heavier element, the first stars switched on.

What they were, how they died, and how we are trying to find one.

Origin

A universe made of two elements

The Big Bang made hydrogen, helium and a trace of lithium. Nothing else.

No carbon, no oxygen, no iron, no dust. Every heavier atom in your body was manufactured later, inside stars, and the first stars had none of it to work with. Astronomers call all those heavier elements metals, and the first stars had a metallicity of exactly zero.

Roughly 100 to 200 million years after the Big Bang, gas drained into the gravitational wells of dark matter clumps around a hundred thousand times the mass of the Sun. In the densest of these minihaloes, hydrogen finally grew cold and dense enough to collapse, and the universe lit up for the first time.

Cosmic history

Where they sit in time

  1. 0Big Bang
  2. 380,000 yrFirst atoms form, the afterglow we see as the CMB
  3. ~100 MyrThe dark ages: gas, dark matter, no light sources
  4. ~150 MyrThe first stars, redshift z ≈ 30–20
  5. ~400 MyrFirst galaxies assemble, z ≈ 10
  6. ~1 GyrReionization completes, z ≈ 6
  7. 13.8 GyrToday

Ages are approximate. Redshift z measures how much the universe has stretched since light left: z = 20 means the universe was 21 times smaller than it is now.

Why they grew so big

A cloud can only collapse if it can cool

Squeeze a gas cloud and it heats up, and heat pushes back. To keep collapsing, the gas has to radiate that warmth away.

A modern cloud is laced with carbon, oxygen and dust, which radiate efficiently and let the gas fall to about 10 kelvin. Pristine gas has almost nothing to work with: only molecular hydrogen, a feeble coolant that is easily broken apart, and which stalls near 200 kelvin.

Warmer gas resists gravity better, so the smallest lump that can collapse under its own weight is far heavier. Simulations of pristine clouds make stars of tens to hundreds of solar masses, often in small multiple systems, instead of the swarm of modest stars a present-day cloud produces.

How heavy they really got is still open. Fragmentation, rotation and radiation from the growing star all fight back, and different simulations land on different mass distributions. It is one of the reasons finding a real Pop III star matters.

Pristine gascools to ~200 K, fragments into a few very massive stars
Enriched gascools to ~10 K, fragments into many small stars

Anatomy

Hotter, bluer, brighter, briefer

Metals make a star opaque. Without them the interior is more transparent, so a first star settles into a more compact, hotter configuration than a modern star of the same weight. A hundred solar masses of pristine hydrogen runs at roughly 100,000 kelvin at the surface, close to twenty times the temperature of the Sun, and radiates most of its light in the ultraviolet.

  • The Sun, for scale
  • Population III star
Radius
—
Surface temperature
—
Luminosity
—
Lifetime
—

Approximate zero-metallicity main-sequence values, interpolated from published stellar models. The Sun is drawn to the same scale for comparison.

Their light

Strong hydrogen, absent oxygen

All that ultraviolet light does not escape quietly. It tears electrons off the surrounding hydrogen, and when the electrons find their nuclei again the gas glows in a ladder of hydrogen lines, brightest among them Hα, deep in the red.

A first star is hot enough to do something ordinary stars cannot do in quantity: strip both electrons from helium. The recombination line that follows, He II at 1640 ångström, is the classic Pop III calling card.

What is missing matters just as much. The line that dominates most young star-forming galaxies, [O III] at 5007 ångström, needs oxygen. A genuinely metal-free nebula has none to give.

Schematic, for the idea rather than the detail: a metal-free nebula (top) against a typical enriched star-forming galaxy (bottom).

Bright hydrogen with no oxygen beside it. That contrast is the fingerprint modern searches are built on.

Populations, not single stars

What a first galaxy would look like

One first star on its own is far beyond the reach of any telescope we have. A cluster of them is not. When a pocket of pristine gas makes hundreds or thousands of stars at once, the result is a small, intensely blue system whose light is dominated by the glowing gas around it.

A Population III galaxy would be tiny by present-day standards, perhaps a few thousand to a few million solar masses of stars inside a dark matter halo a hundred times heavier. Its emission lines would be extraordinarily strong, carrying a large share of the total light, and the He II line would stand there without a single metal line beside it.

Completely metal-free galaxies are expected to be rare and short-lived, because the first supernova pollutes its own surroundings. The more realistic target is a pristine pocket: a metal-free clump forming stars at the edge of, or just beside, a galaxy that has already been enriched. Several of the candidates reported so far look like this, and magnification by a foreground cluster of galaxies is usually what brings them within reach.

Death

Four ways to end

A first star lives for a few million years at most, then its fate is decided almost entirely by its mass. The boundaries below come from stellar models and shift with rotation and binarity, but the pattern is robust.

Legacy

They are gone, but they are not lost

The first metals

One Pop III supernova can seed its whole halo. Once the gas carries even a thousandth of the Sun's metals, cooling becomes easy and star formation switches to the smaller, longer-lived stars we see everywhere today.

Fossil chemistry

The oldest stars in the Milky Way's halo are made of that polluted gas. The most primitive of them, carbon-rich and iron-poor, carry abundance patterns that look like the yield of a single massive first-generation supernova.

Black hole seeds

Quasars weighing a billion Suns are already in place less than a billion years after the Big Bang. Massive first stars collapsing directly into black holes are one of the few ways to start something that heavy that early.

Clearing the fog

Their ultraviolet output began ionizing the neutral hydrogen between galaxies, the long process that ends with the transparent universe we look through now.

The end

It ended in patches, not all at once

Enrichment is a local business. A supernova pollutes its own neighbourhood within a few million years, but gas sitting far from any explosion can stay pristine for hundreds of millions of years longer.

Most models put the peak of metal-free star formation somewhere around 200 to 400 million years after the Big Bang, redshift 20 to 10, followed by a slow decline as metals spread outwards from the first galaxies.

That decline has a long tail. Simulations keep making a trickle of Population III stars down to redshift 6, and in several of them below redshift 5 as well, long after most galaxies had settled into ordinary, metal-rich star formation. If that is right, the last metal-free stars may have formed when the universe was already billions of years old, in remote pockets that never met a supernova.

The tail is what makes the search realistic. The first stars at redshift 20 are hopelessly out of reach, while a pristine pocket at redshift 6 to 8 sits comfortably inside what JWST can see.

How long the tail lasts depends on how efficiently metals mix into fresh gas, which is one of the least certain parts of the whole picture.

Schematic: metal-free star formation peaks early and then trails off rather than stopping, with a low rate persisting past redshift 5 in many models. Only the right-hand end of the curve is within reach of current telescopes.

Open questions

What we still do not know

Almost everything on this page rests on theory and simulation, because nobody has yet seen a first star. These are the pieces most likely to change once somebody does.

How heavy were they?

The earliest simulations made single stars of several hundred solar masses. Newer ones, which follow the disc of gas around the growing star, often break it into several smaller pieces. The mass distribution of the first stars, the number that almost everything else depends on, is still unsettled.

Could any still be alive?

A metal-free star below roughly eight tenths of a solar mass would outlive the present age of the universe, and could be sitting in the Milky Way's halo right now. Surveys have turned up spectacularly metal-poor stars, but none with a metallicity of zero. Either such stars never formed, or they are very rare, or their surfaces have been dirtied since.

How fast did they spin?

Rapid rotation mixes fresh elements up to the surface, changes how much ultraviolet light escapes, and can turn a collapse into a long gamma-ray burst. The spin rates of the first stars are essentially unconstrained.

Were they usually in pairs?

If massive first stars commonly formed in binaries, the black holes they left behind could spiral together and merge. Whether any of the mergers detected in gravitational waves began this way is an open question.

Did they seed the giant black holes?

Quasars weighing a billion Suns are in place less than a billion years after the Big Bang. Whether their seeds were collapsing first stars, direct collapse of pristine gas clouds, or something else, is unresolved.

Is any candidate real?

Several systems show hints, usually helium emission with no metal lines beside it. Confirming one takes deep spectroscopy and the patient exclusion of more ordinary explanations, and so far none has survived that test.

Go deeper

Further reading