Where the elements come from

Foundations

The furnace that closed too early

Lesson 1.2 left the universe with a fixed recipe: three quarters hydrogen, one quarter helium, a trace of lithium, and nothing else. That was twenty minutes after the beginning, and the furnace had gone out. There was no carbon, no oxygen, no iron, no gold. There was nothing to make a planet from, let alone a person.

Everything else came later, and it came from stars. Not as a by-product of their shining but as the shining itself: the light of every star is the exhaust of a factory turning light elements into heavy ones. This chapter is about that factory.

Hoyle’s problem

In 1946 Fred Hoyle asked how a star could build the periodic table, and immediately hit a wall. To get past helium you must combine helium nuclei, and the obvious step, two heliums making beryllium-8, produces a nucleus so unstable it falls apart in a ten-millionth of a billionth of a second. To make carbon, a third helium nucleus must arrive within that time. Edwin Salpeter showed in 1952 that in a red giant’s core, at a hundred million degrees and enormous density, a tiny equilibrium population of beryllium-8 exists at any moment, and occasionally a third helium finds one. Three heliums make carbon. Astronomers call it the triple-alpha process.

The rate still came out too slow to make the carbon we see. So Hoyle did something audacious. He reasoned that since carbon plainly exists, and since we exist and are made of it, the carbon nucleus must have an excited state at just the right energy to make the reaction resonate, near 7.65 million electron volts. He walked into a laboratory at Caltech in 1953 and asked them to look for it. They looked. It was there, within a fraction of a percent of where he said it would be.

That state is why the universe has chemistry. Without it, stars would make helium and stop, and the cosmos would be a place of gas and light with nothing solid in it.

Up the ladder

Once a star has carbon, the rest follows. Add a helium nucleus to carbon and you get oxygen; add another and you get neon, then magnesium, then silicon, sulphur, argon, calcium. This is why the even-numbered elements are so much more common than their odd-numbered neighbours. The universe builds in steps of two protons because helium nuclei are the building blocks lying around.

Each stage needs a hotter core, and only a massive star ever gets there. Carbon burning at six hundred million degrees lasts centuries; neon burning about a year; oxygen burning a few months; silicon burning about a day. By the end, the star is an onion of shells, each fusing a different element, wrapped around a core of iron.

Why it stops at iron

Iron is not special because of anything chemical. It is special because of how tightly its nucleus is bound. Plot how firmly the particles in a nucleus are held together against the size of the nucleus, and the curve rises steeply from hydrogen, flattens, and peaks around iron and nickel, then declines slowly all the way to uranium.

Fusing anything below the peak releases energy, which is why stars shine. Fusing anything above it costs energy. When a star’s core becomes iron, the furnace has reached the bottom of the valley, and there is nothing left to burn. The core collapses, which is the supernova of lesson 2.3 and the subject of the next lesson.

The gentle route past iron

Not everything heavier than iron waits for an explosion. In the outer layers of ageing giant stars, reactions release free neutrons, and neutrons have no charge, so they can drift into a nucleus without fighting the electrical repulsion that makes fusion so hard. A nucleus captures a neutron every few thousand years, has plenty of time to settle or decay in between, and creeps slowly up the periodic table. This slow neutron capture builds about half the elements above iron: strontium, barium, lead and much else. The star then dredges them to the surface and blows them away in its wind, which is why the spectra of some old giants show elements that could only have been made inside them.

Reading the receipts

This is a story about events that happened billions of years ago inside objects nobody can visit, so it needs evidence.

The first is the pattern of abundances. Measure how much of every element the solar system contains, from the Sun’s spectrum and from meteorites, and the resulting curve is jagged in a very particular way: the even-odd sawtooth, the peak at iron, the twin bumps above it from slow and fast neutron capture. Every feature matches a specific process.

The second is more direct. Some meteorites contain microscopic grains of silicon carbide and graphite and diamond that condensed in the winds of dying stars before the Sun existed, survived the building of the solar system, and can be picked out with tweezers and a microscope. Their isotope ratios are wildly different from anything in the solar system, and they match what nucleosynthesis calculations predict for the stars that made them. They are physical samples of other stars’ output, sitting in laboratory drawers.

The elements that fit nowhere

Three light elements, lithium, beryllium and boron, sit awkwardly. Stars destroy them rather than making them, and the Big Bang made only a little lithium. Most of what exists was made in interstellar space, when cosmic rays travelling near the speed of light smashed into carbon, nitrogen and oxygen nuclei and chipped them into fragments. It is the one part of the periodic table built by demolition rather than construction.

Deeper

Triple alpha in numbers

Two alpha particles form beryllium-8, which is unbound by 92 keV and decays with a lifetime near 10⁻¹⁶ seconds. At 10⁸ kelvin and densities around 10⁵ grams per cubic centimetre, the equilibrium abundance of beryllium-8 is roughly one part in 10⁹ of helium: small, but not zero. The capture of a third alpha proceeds through the 0⁺ excited state of carbon-12 at 7.654 MeV, just 288 keV above the beryllium-8 plus alpha threshold, which makes the reaction resonant at helium-burning temperatures. The rate depends on the cube of the helium density and, near 10⁸ kelvin, on temperature to roughly the fortieth power. About one in every 2,400 carbon-12 nuclei formed in this state decays to the ground state rather than back to three alphas, and that leak is the source of all the carbon there is.

The burning stages

Each stage produces the fuel for the next and leaves an unburnt shell behind, so the star develops the onion structure. The timescales shorten because neutrino losses from the hot core grow enormously with temperature, draining energy far faster than photons can carry it away.

The binding-energy curve

Binding energy per nucleon rises from 1.1 MeV for deuterium to about 8.8 MeV in the iron region and falls to 7.6 MeV for uranium. The maximum belongs to nickel-62 at 8.7945 MeV per nucleon, with iron-58 and iron-56 just below; iron-56 dominates in stellar cores because silicon burning proceeds through nuclear statistical equilibrium, which at the relevant proton-to-neutron ratio favours nickel-56, and nickel-56 then decays to cobalt-56 and iron-56. The radioactive decay of that nickel-56 is what powers the visible light curve of a supernova for weeks afterwards.

The s-process

In helium-burning shells of asymptotic giant branch stars, the reactions carbon-13 with alpha, and neon-22 with alpha, release free neutrons at densities of 10⁷ to 10¹¹ per cubic centimetre. Neutron capture then proceeds slowly relative to beta decay, so the reaction path hugs the valley of stability, climbing from iron seeds up to lead and bismuth over tens of thousands of years. Bottlenecks occur at nuclei with closed neutron shells, at mass numbers 88, 138 and 208, which is why strontium, barium and lead have abundance peaks. Convection dredges the products to the surface, and technetium, whose longest-lived isotope survives only a few million years, has been seen in the spectra of such stars: proof that nucleosynthesis is happening now, in stars we can observe.

The abundance curve as evidence

The solar system abundance pattern, compiled by Suess and Urey in 1956 and refined ever since, shows hydrogen and helium dominating, a deep trough at lithium, beryllium and boron, a broad peak in the carbon-to-calcium range, a sharp maximum at iron, and a long tail with paired peaks at the s-process and r-process closed shells. B²FH’s achievement was to assign each feature to a distinct physical process. No feature is unexplained, and the quantitative agreement between measured abundances and computed yields is the strongest evidence that the theory is right.

Frontier

How much escapes

Calculating what a massive star makes is now routine; calculating how much of it gets out is not. Material near the collapsing core may fall back onto the neutron star or black hole rather than escaping, and the amount depends on the explosion mechanism, which lesson 2.3 noted is still not fully simulated. Predicted yields of iron-peak elements vary by factors of a few between models. Comparing predictions with the compositions of the oldest, most metal-poor stars, each of which was enriched by only one or a few earlier stars, is one way of testing them.

Reaction rates in the dark

Several key reactions, above all carbon-12 plus helium making oxygen-16, are still not measured precisely at stellar energies, because the rates are far too low to measure directly in a laboratory and must be extrapolated. That single rate sets the ratio of carbon to oxygen in the universe, and thus in every living thing. Underground accelerators are chipping away at it.

Transcript and shot list

Target runtime about 7 minutes. Built as follow the build order: start at the wall that nearly stopped everything, then climb the table one fusion stage at a time until the furnace reaches iron and stops. Shot types: ANIM is rendered from code with Manim, AI is a generated shot from Higgsfield with the shared style reference, REAL is agency imagery with credit, CAM is the narrator.

[00:00] S01 · AI · A hand holding an iron nail; behind it, a red giant star fills the sky.

Here’s an iron nail. Every atom in it was made inside a star, in the last few days of that star’s life, at two and a half billion degrees. Nothing on Earth has ever been that hot. This chapter is about how the universe went from three ingredients to ninety-two, and it starts with a problem that nearly stopped everything.

[00:28] S02 · ANIM · Cosmic calendar; a marker at 8 January labelled “first heavy elements”; a bar running to now.

Around the 8th of January on the cosmic calendar, the first massive stars died and scattered the first heavy elements. The factory has been running ever since.

[00:40] S03 · ANIM · The recipe from lesson 1.2: hydrogen, helium, a trace of lithium; a periodic table with only the first three squares filled and the rest grey.

Lesson 1.2 left the universe with a fixed recipe. Three quarters hydrogen. One quarter helium. A trace of lithium. And that’s all, because the furnace went out after twenty minutes. No carbon. No oxygen. No iron. Nothing to build a planet from, let alone a person.

[01:01] S04 · ANIM · Two helium nuclei fuse into beryllium-8, which flies apart; a clock shows 10⁻¹⁶ seconds; repeat, faster; a wall appears.

In 1946 Fred Hoyle asked how a star could get past helium, and hit a wall. Combine two helium nuclei and you get beryllium-8, which falls apart in a ten-millionth of a billionth of a second. To make carbon, a third helium has to arrive inside that window. It sounds impossible.

[01:25] S05 · ANIM · A red giant’s core at 10⁸ K; a small standing population of beryllium-8; a third helium arrives; carbon forms; label: triple-alpha.

Edwin Salpeter showed in 1952 that it isn’t. In a red giant’s core, at a hundred million degrees and enormous density, there’s always a tiny standing population of beryllium-8. Now and then a third helium finds one before it decays. Three heliums, one carbon.

[01:45] S06 · ANIM · The rate is still too slow; then a resonance level appears in the carbon-12 nucleus at 7.65 MeV and the rate leaps; typographic: Caltech, 1953.

But the rate came out far too slow to make the carbon we actually see. So Hoyle did something audacious. He reasoned that carbon plainly exists, and we’re made of it, so the carbon nucleus must have an excited state at just the right energy to make this reaction resonate. About seven point six five million electron volts. He walked into a lab at Caltech in 1953 and asked them to look. They looked. It was there, within a fraction of a percent of where he said. That energy level is why the universe has chemistry at all. Without it, stars would make helium and stop.

[02:34] S07 · ANIM · Helium nuclei added one at a time: carbon, oxygen, neon, magnesium, silicon; an abundance chart shows even-numbered elements standing taller than odd.

Once a star has carbon, the rest follows. Add a helium nucleus and you get oxygen. Add another: neon. Then magnesium, silicon, sulphur, argon, calcium. That’s why the even-numbered elements are so much more common than their odd neighbours. The universe builds in steps of two protons, because helium nuclei are what’s lying around.

[02:59] S08 · ANIM · The onion shells of a massive star; a table of stage durations: centuries, a year, months, one day; the core turns to iron.

Each stage needs a hotter core, and only a massive star gets there. Carbon burning lasts centuries. Neon, about a year. Oxygen, months. Silicon, one day. By the end the star is an onion of burning shells around a core of iron.

[03:18] S09 · ANIM · The binding-energy curve rising to a peak at iron and falling; arrows showing fusion releasing energy below the peak and costing energy above.

And here’s why it stops. Plot how tightly the particles in a nucleus are held, against the size of the nucleus. The curve rises steeply, flattens, peaks around iron and nickel, then falls. Fusing anything below the peak releases energy. That’s why stars shine. Fusing anything above it costs energy. When the core is iron, the furnace has reached the bottom of the valley. There’s nothing left to burn, and the star has seconds to live.

[03:54] S10 · ANIM · In an ageing giant’s shell, free neutrons drift into nuclei; a slow climb up the periodic table over thousands of years; strontium, barium, lead light up.

Not everything heavier than iron waits for the explosion. In the outer layers of ageing giants, reactions release free neutrons, and a neutron has no charge, so it can drift into a nucleus without fighting any repulsion. One capture every few thousand years, with plenty of time to settle in between. Slowly, a nucleus creeps up the table. This builds about half the elements above iron: strontium, barium, lead. The star then dredges them up and blows them away in its wind.

[04:31] S11 · ANIM · The solar abundance curve with its sawtooth, iron peak and twin bumps; each feature labelled with its process. S12 · REAL · A presolar silicon carbide grain under an electron microscope (NASA or a laboratory image; confirm credit line).

How could anyone check this? Two ways. First, the pattern. Measure how much of every element the solar system holds, and the curve is jagged in a very particular way: the even-odd sawtooth, the peak at iron, the twin bumps above it. Every feature matches a specific process. Second, and better: some meteorites contain microscopic grains that condensed around dying stars before the Sun existed. They survived the making of the solar system. You can pick them out with tweezers, and their isotopes match what the calculations predict for the stars that made them. Those are physical samples of other stars, sitting in laboratory drawers.

[05:20] S13 · ANIM · Lithium, beryllium and boron highlighted; a cosmic ray smashes a carbon nucleus into fragments.

Three elements sit awkwardly. Lithium, beryllium and boron: stars destroy them rather than making them. Most of what exists was made in empty space, when cosmic rays travelling near light speed smashed into carbon and oxygen nuclei and chipped them apart. One corner of the periodic table built by demolition.

[05:43] S14 · ANIM · Labels: ESTABLISHED for the processes and the abundance match; ACTIVE RESEARCH for yields and the carbon-oxygen reaction rate.

The scorecard. That stars build the elements from carbon to iron, that slow neutron capture builds half of what’s above iron, and that the measured abundances match: established since 1957, and confirmed by dust grains you can hold. How much of each element actually escapes a dying star, and one crucial reaction rate that sets the ratio of carbon to oxygen in the universe: active research.

[06:13] S15 · AI into ANIM · The iron nail again; it dissolves into the onion shells of a dying star; the calendar returns.

So, the nail. Iron, from the last day of a massive star. But the star had to explode for it to reach you, and that explosion is where the story gets violent. It’s also where gold comes from. Next lesson.

Shot list

Two AI shots, twelve code-rendered shots, one microscope image. Estimated Higgsfield use: 3 to 6 generations including retries.