What Is Nucleosynthesis? How Atoms Are Forged in Stars and Stellar Explosions

"We are made of star stuff." Carl Sagan's famous statement is not metaphor — it's astrophysics. The atoms in your body were forged in one or more of the universe's great nucleosynthetic factories: in the Big Bang itself, in the cores of long-dead stars, or in the catastrophic explosions of those stars' deaths. Some of the gold in human civilisation's jewellery was made in a collision between two neutron stars that occurred before the solar system existed.

Nucleosynthesis — the creation of atomic nuclei from protons, neutrons, and pre-existing nuclei — is the process that built the periodic table of elements from the universe's original near-pure hydrogen. Understanding where each element comes from reveals a cosmic history of fire, gravity, and violence written in the atoms of everything that exists.

The problem: stars can't make everything

Before nucleosynthesis was understood, astronomers faced a puzzle: stars contain elements far heavier than hydrogen and helium, but how could those elements be made inside stars?

The answer, developed primarily by Fred Hoyle, William Fowler, and Geoffrey and Margaret Burbidge in the landmark 1957 paper "Synthesis of the Elements in Stars", known as B²FH after its authors' initials (published in Reviews of Modern Physics, volume 29, page 547), described multiple distinct nucleosynthetic processes, each operating in different conditions. Their work (along with earlier contributions from Hans Bethe, who described the proton-proton chain and the CNO cycle in the late 1930s, and an independent 1957 treatment by Alastair Cameron) established the field of nuclear astrophysics.

The basic insight: different elements require different physical conditions — different temperatures, densities, neutron fluxes — and those conditions are found in different astrophysical environments. No single process makes all elements; the periodic table is assembled from contributions from many cosmic sources.

How we learned what stars are made of

The story of nucleosynthesis starts with a question that sounds simple: what are stars made of? In the nineteenth century, the spectroscope showed that each element absorbs light at its own set of wavelengths, so the dark lines in a star's spectrum work like a chemical barcode. Astronomers could now identify iron, calcium, sodium and hydrogen in sunlight without leaving Earth.

What they could not easily do was turn line strengths into amounts. In her 1925 doctoral thesis, Cecilia Payne applied the new physics of ionisation to stellar spectra and concluded that stars are made overwhelmingly of hydrogen and helium. The result was so far from the expectation of the time, that the Sun had roughly Earth-like composition, that she was persuaded to describe it as probably not real. Within a few years, other astronomers had confirmed it. Modern measurements give the Sun's surface composition as roughly three-quarters hydrogen and one-quarter helium by mass, with everything else adding up to only about 1–2%.

That left a two-part puzzle. Where did all the hydrogen and helium come from, and how did the small but crucial fraction of heavier elements get made? In 1948, Ralph Alpher, Hans Bethe and George Gamow published a short paper proposing that the elements were built by neutron capture in a hot, dense early universe. They were right about helium but wrong about the rest: the mass gaps at 5 and 8 meant the early universe could not build much beyond lithium. Heavier elements needed a different furnace, and B²FH showed that stars, and their deaths, were it.

Two observations made the stellar picture convincing. In 1952, the astronomer Paul Merrill detected technetium in the spectra of certain red giant stars. Technetium has no stable isotopes, and its longest-lived forms decay over a few million years at most, far shorter than these stars' ages. It could only be there if the stars were making it themselves. And in 1953, a team at Caltech found the excited state of carbon-12 that Hoyle had predicted had to exist for stars to make carbon (see below). The physicist William Fowler, who worked on that measurement and spent decades measuring nuclear reaction rates, shared the 1983 Nobel Prize in Physics for this work.

Big Bang nucleosynthesis (BBN): the universe's first chemistry

When: The first 10 seconds to 20 minutes after the Big Bang Temperature range: ~10¹² K to ~10⁸ K Products: Hydrogen (¹H), deuterium (²H), helium-3 (³He), helium-4 (⁴He), lithium-7 (⁷Li), and trace amounts of beryllium-7 (⁷Be, which decays to ⁷Li)

In the universe's first second, temperatures were so extreme that protons and neutrons were constantly forming and transforming — the universe was an undifferentiated plasma where particles and antiparticles annihilated each other. As the universe expanded and cooled to ~10¹² K (around 1 second), protons and neutrons stopped converting into each other freely. The ratio froze at approximately 7 protons to 1 neutron.

As temperatures dropped further, protons and neutrons began binding. Deuterium (one proton + one neutron) formed but was immediately destroyed by high-energy photons. Finally, at ~10⁹ K (about 100 seconds), the photon density dropped enough that deuterium could survive — and a rapid sequence of nuclear reactions commenced:

• D + D → ³He + n
• D + D → ³H + p
• ³He + D → ⁴He + p
• ³H + D → ⁴He + n
• ³He + n → ³H + p

The end result: virtually all available neutrons were incorporated into helium-4 (the most stable light nucleus). The final composition: approximately 75% hydrogen, 25% helium-4 by mass, with traces of deuterium (about 2.5 deuterium atoms for every 100,000 hydrogen atoms), helium-3, and lithium-7.

After about 20 minutes, the universe cooled enough that nuclear reactions effectively ceased. No heavier elements were produced — the density and temperature conditions lasted too briefly for nuclei heavier than beryllium to build up significantly. (A famous "mass gap" at mass-5 and mass-8 — where no stable nuclei exist — blocked further synthesis during BBN.)

The predictive power of BBN: The exact abundances predicted by BBN theory depend on a single parameter: the baryon density of the universe (the density of protons and neutrons). The observed deuterium abundance in primitive gas clouds at high redshift, measured spectroscopically, matches the BBN prediction. This is one of the most precise tests of the standard cosmological model — the observed deuterium-to-hydrogen ratio, about 2.51 × 10⁻⁵ in the Particle Data Group average, sits within roughly 3% of the standard prediction (about 2.44 × 10⁻⁵).

For the physics of the first few minutes before nucleosynthesis began, see our explainer on how the Big Bang actually happened.

The lithium problem: There is a persistent tension: metal-poor halo stars (old, primordial-composition stars) show lithium abundances about three to four times lower than BBN predicts (observed ⁷Li/H around 1.5 × 10⁻¹⁰ versus a predicted value near 5.5 × 10⁻¹⁰). This "cosmological lithium problem" could reflect stellar depletion of lithium over time, or subtle errors in BBN modelling. It remains unresolved.

Stellar nucleosynthesis: building the middle of the periodic table

Everything from carbon to iron is primarily made inside stars through stellar nucleosynthesis:

Hydrogen burning: helium synthesis

The main-sequence life of a star is one long nucleosynthesis reaction (our article on how the Sun works covers the energy side of this story) — converting hydrogen to helium. Over 10 billion years, the Sun will convert approximately 10% of its mass from hydrogen to helium.

The p-p chain (dominant in the Sun):

1. ¹H + ¹H → ²H + e⁺ + ν (the slowest step — takes ~10¹⁰ years on average for an individual proton pair in the Sun's core)
2. ²H + ¹H → ³He + γ
3. ³He + ³He → ⁴He + 2¹H

Net reaction: 4 ¹H → ⁴He + 2e⁺ + 2ν + energy

Energy released per reaction: 26.73 MeV (4.28 × 10⁻¹² joules) — modest, but occurring 9.2 × 10³⁷ times per second in the Sun's core.

Helium burning: carbon and oxygen

When core hydrogen is exhausted, helium burning begins at ~10⁸ K:

The triple-alpha process (3α): Three helium nuclei (alpha particles) fuse to form carbon-12. This is more complex than simple two-body reactions — it requires the temporary formation of beryllium-8 (which has a lifetime of only 10⁻¹⁶ seconds), followed by rapid capture of a third helium nucleus.

The triple-alpha rate depends sensitively on the energy of an excited state of carbon-12 (the "Hoyle state" — predicted by Fred Hoyle before its experimental discovery, in one of the most successful theoretical predictions in nuclear physics). Without this energy level (7.65 MeV above ground state), the triple-alpha rate would be far too slow and carbon wouldn't accumulate in stars.

After carbon forms: alpha capture continues

• ¹²C + ⁴He → ¹⁶O + γ
• ¹⁶O + ⁴He → ²⁰Ne + γ (less efficient at typical helium-burning temperatures)

The ratio of carbon to oxygen produced depends on the uncertain rate of the ¹²C + α reaction — one of nuclear physics' most important unmeasured quantities, affecting all subsequent stellar nucleosynthesis.

Carbon burning (~10⁸·⁷ K, in massive stars only)

• ¹²C + ¹²C → ²⁴Mg + γ
• ¹²C + ¹²C → ²³Na + ¹H
• ¹²C + ¹²C → ²⁰Ne + ⁴He
• ¹²C + ¹²C → ²³Mg + n

Products: neon, sodium, magnesium

Neon burning (~1.5 × 10⁹ K)

Photodisintegration (high-energy photons breaking apart neon) releases helium nuclei, which are captured by remaining neon:

• ²⁰Ne + γ → ¹⁶O + ⁴He
• ²⁰Ne + ⁴He → ²⁴Mg + γ

Products: oxygen, magnesium

Oxygen burning (~2 × 10⁹ K)

• ¹⁶O + ¹⁶O → ³²S + γ
• ¹⁶O + ¹⁶O → ³¹P + ¹H
• ¹⁶O + ¹⁶O → ²⁸Si + ⁴He
• ¹⁶O + ¹⁶O → ³¹S + n

Products: silicon, sulfur, phosphorus, chlorine, argon, potassium

Silicon burning (~3 × 10⁹ K)

At these temperatures, photodisintegration reactions are so intense that silicon and its products are in a near-nuclear statistical equilibrium (NSE) — nuclei are constantly forming and breaking apart, and the equilibrium favours iron-peak elements: iron (⁵⁶Fe), nickel (⁵⁶Ni), chromium, cobalt, manganese.

Silicon burning produces the iron core that will trigger core collapse. It's not a simple "silicon burns to iron" process but rather a complex network of reactions at quasi-statistical equilibrium.

The iron wall: Nuclei around iron and nickel sit at the peak of the binding-energy curve. Strictly, nickel-62 has the highest binding energy per nucleon (about 8.79 MeV), with iron-58 and iron-56 just behind; iron-56 has the lowest mass per nucleon. The popular line that "iron-56 is the most stable nucleus" blurs those two measures. What matters for stars is that fusing nuclei beyond this peak absorbs energy rather than releasing it. Silicon burning actually makes mostly nickel-56, which later decays through cobalt-56 to iron-56, and that is why iron is so abundant. Once the core is iron-peak material, no more energy can be extracted by fusion, and the star's support collapses.

Neutron capture: building heavy elements

Elements beyond the iron peak (elements 27–92+) cannot be made by charged-particle fusion — the Coulomb barrier becomes too strong for fusion to overcome at temperatures found in stars. Instead, they're built by neutron capture — neutrons, lacking charge, can be captured by nuclei at any energy.

Two distinct neutron capture processes operate in different environments:

The s-process (slow neutron capture)

Where: Interior of AGB stars (thermally pulsing, low-to-intermediate mass stars in late stages) Timescale: Thousands to millions of years

The s-process occurs when a seed nucleus (typically iron) captures neutrons slowly — one at a time, with time between captures for beta decay (n → p + e⁻ + ν̄) to occur. The nucleus climbs the periodic table in a predictable path that follows the "valley of stability" — because unstable isotopes have time to decay before the next capture.

The neutron flux in AGB stars comes from ¹³C + ⁴He → ¹⁶O + n (the "¹³C pocket") and ²²Ne + ⁴He → ²⁵Mg + n reactions during thermal pulses.

s-process products: About half of all elements between barium and lead — including barium, strontium, yttrium, zirconium, lanthanum, cerium, and lead. The s-process terminates at lead-208 (or bismuth-209), where neutron capture eventually leads to unstable isotopes that alpha-decay back down.

AGB stars return this s-process-enriched material to the interstellar medium through their intense stellar winds.

The r-process (rapid neutron capture)

Where: supernova. About 20 km wide but containing more mass than the Sun.">Neutron star mergers (kilonovae); possibly also rare "collapsar" supernovae; possibly magnetically driven supernovae Timescale: Fractions of a second to seconds

The r-process occurs under conditions of extreme neutron flux — so intense that a nucleus captures many neutrons before it has time to decay. Nuclei are pushed far from the valley of stability into the neutron-rich region. When the neutron flux ends (after the explosion or merger), the neutron-rich nuclei cascade back to stability through rapid beta decays.

r-process products: The remaining ~50% of elements beyond iron not made by the s-process, including thorium, uranium, plutonium, and notably gold (⁷⁹Au) and platinum (⁷⁸Pt).

The smoking gun: The 2017 neutron star merger event GW170817 — the first multi-messenger astronomical event observed in both gravitational waves (LIGO) and electromagnetic radiation — provided direct observational evidence for r-process nucleosynthesis.

The kilonova (electromagnetic transient from the merger ejecta) showed spectroscopic signatures characteristic of heavy r-process elements — including strontium (confirmed) and signatures consistent with gold and platinum. Modelling of the light curve by Daniel Kasen and colleagues in 2017 found two distinct ejecta components: one dominated by lighter r-process nuclei (mass number below about 140) and one by heavy ones (above 140), and argued that mergers like this one could be the dominant source of r-process elements in the universe. Published estimates of the total ejecta are a few hundredths of a solar mass, which is thousands of Earth masses of freshly made heavy elements. Estimates of the gold alone differ widely between studies, so treat any single figure with caution.

In 2019, Darach Watson and colleagues reanalysed the kilonova's spectra from ESO's Very Large Telescope and identified strontium, the first individual neutron-capture element pinned down in a merger.

The meteorite record adds a local story. In a 2019 Nature paper, Imre Bartos and Szabolcs Márka compared models with the abundances of short-lived actinides, such as plutonium-244 and curium-247, in early solar system material. They concluded that a single neutron-star merger about 300 parsecs (roughly 1,000 light-years) away, about 80 million years before the solar system formed, probably supplied much of the early solar system's curium and a substantial fraction of its plutonium.

Who makes what: comparing the cosmic factories

Each source in the cosmic recipe has a different signature. Their relative importance is still being refined, but the broad pattern is well established.

Source	Typical conditions	Main products	How long after a galaxy forms it starts contributing
Big Bang	About 10⁹ K, first ~20 minutes	¹H, ²H, ³He, ⁴He, ⁷Li	Before any galaxy existed
Massive stars and core-collapse supernovae	Cores up to several billion K	Oxygen, neon, magnesium, silicon, sulfur, calcium, some iron	Millions of years (massive stars live briefly)
AGB stars (low and intermediate mass)	Helium-shell flashes, modest neutron flux	Carbon, nitrogen, s-process elements such as barium and lead	Hundreds of millions to billions of years
Type Ia supernovae (exploding white dwarfs)	Thermonuclear burning of carbon–oxygen	Most of the iron-peak elements, especially iron and manganese	Typically a billion years or more
Neutron star mergers	Extreme neutron densities, seconds	Heavy r-process elements: europium, gold, platinum, uranium	Delayed by the time binaries take to spiral together
Cosmic rays	High-energy collisions in interstellar space	⁶Li, beryllium, boron	Continuous

The timing column matters because it lets astronomers read a galaxy's history from its stars. The oldest stars in the Milky Way's halo have more oxygen and magnesium relative to iron than the Sun does. That is the fingerprint of early enrichment by massive stars and core-collapse supernovae, before Type Ia supernovae had time to add their iron. Our guide to the different types of supernova explains why white dwarf explosions and core collapses produce such different mixtures.

There is also a live debate over whether neutron-star mergers can explain all of the r-process. Some very old stars are rich in r-process elements, which is hard to explain if mergers always arrive after a long delay. That is why rarer events, such as collapsars (the collapse of rapidly rotating massive stars into black holes) and magnetically driven supernovae, remain on the list of candidate r-process sites. The 2017 merger, detected through gravitational waves, proved that mergers make r-process elements. Whether they make nearly all of them is still open.

The p-process: proton-rich isotopes

A small fraction of isotopes are proton-rich — they can't be made by neutron capture because they lie on the proton-rich side of the stability valley. These are the p-nuclei, made by the p-process (or γ-process):

During core-collapse supernovae, high-energy photons (gamma rays) can cause photodisintegration of neutron-rich nuclei — knocking off neutrons to produce proton-rich species. Alternatively, proton capture at high temperatures can build proton-rich nuclei.

P-process elements include molybdenum-92, -94; ruthenium-96, -98; and ~35 other rare isotopes. They're the least abundant isotopes of their elements — produced in small quantities in events that are rare but energetic.

The vp-process and νp-process: in the supernova wind

In the neutrino-driven wind immediately above a newly formed neutron star, conditions may enable the νp-process — neutrinos converting neutrons to protons, enabling proton capture on nuclei in a neutron-poor environment. This may produce some proton-rich nuclei in the molybdenum-through-tin range.

Cosmic ray spallation: lithium, beryllium, boron

The lightest elements beyond hydrogen and helium — lithium, beryllium, and boron — are destroyed rather than produced in stellar interiors (they're burned at relatively low temperatures). Yet they exist in the universe.

Their source: cosmic ray spallation — high-energy cosmic rays (primarily protons and helium nuclei) striking carbon, nitrogen, and oxygen nuclei in the interstellar medium. The collision shatters the target nucleus, producing lighter fragments including lithium-6, lithium-7, beryllium-9, boron-10, and boron-11.

This explains the high lithium-6/lithium-7 ratio in cosmic rays compared to BBN-produced lithium — ⁶Li is primarily a cosmic ray spallation product, while ⁷Li has both BBN and stellar/cosmic-ray origins.

The origin of every element: a summary

Element range	Primary source
H (hydrogen)	Big Bang nucleosynthesis
⁴He (helium-4)	~25% Big Bang, ~75% stellar H burning
³He, ²H (deuterium)	Big Bang
Li, Be, B	Cosmic ray spallation; partial Big Bang for ⁷Li
C, O	AGB stars, massive stars (He burning)
N	CNO cycle in massive stars; AGB
Ne, Mg, Si, S, Ar, Ca	Massive stars (C, Ne, O burning)
Iron-peak (Cr–Zn)	Core-collapse SNe; Type Ia SNe
Elements 30–82 (½)	s-process in AGB stars
Elements 30–82 (½) + 83–92	r-process in neutron star mergers
Gold, platinum	Primarily neutron star mergers
Uranium, thorium	r-process in neutron star mergers
Technetium, promethium	No stable isotopes; only in active reactors/stars

A personal chemical inventory

Consider some of the elements in a human body (~70 kg adult):

• Oxygen (43 kg): Made in massive stars by helium burning and released in supernovae
• Carbon (16 kg): Made in AGB stars by the triple-alpha process
• Hydrogen (7 kg): Made in the Big Bang, 13.8 billion years ago
• Nitrogen (1.8 kg): Made in CNO cycle of massive stars
• Calcium (1 kg): Made in oxygen burning in massive stars
• Phosphorus (0.5 kg): Made in oxygen burning in massive stars
• Iron (4.2 g): Made in Type Ia and core-collapse supernovae
• Zinc (2.5 g): Mix of s-process and r-process contributions
• Iodine (20 mg): r-process and s-process contributions
• Gold (0.2 mg, trace): Primarily neutron star merger r-process

You are the product of at least three distinct nucleosynthetic processes operating across 13.8 billion years of cosmic history — Big Bang nucleosynthesis, generations of stellar evolution and supernovae, and at least one neutron star merger.

The future of nucleosynthesis

Nucleosynthesis continues today throughout the universe — in every star that is burning, every supernova that explodes, every neutron star that merges. But the overall character of nucleosynthesis is changing:

Metallicity evolution: The universe's average metallicity (heavy element content) increases with each generation of stars. The first stars had zero heavy elements; the Sun formed with a relatively high metallicity; future generations of stars will be even more metal-rich.

Isotopic fingerprints: The isotopic ratios of elements in meteorites — particularly calcium-aluminum-rich inclusions — record the isotopic signatures of specific nucleosynthetic events that occurred in the stellar neighbourhood just before the solar system formed. Anomalous isotope ratios indicate contributions from specific types of supernovae or AGB stars.

The Solar System's nucleosynthetic heritage: The presence of short-lived radioactive isotopes in the early solar system (²⁶Al, ⁶⁰Fe, ¹⁸²Hf) is widely interpreted as pointing to a nearby core-collapse supernova (or similar massive-star source) shortly before or during solar system formation, though the source and timing are still debated — a supernova that some researchers think may have triggered the collapse of the molecular cloud that became the Sun.

The atoms you're made of were assembled across 13.8 billion years in environments ranging from the first three minutes after the Big Bang to the catastrophic merger of two neutron stars. We are not separate from the universe — we are made of the universe, reassembled by evolution.

Common misconceptions

"Stars make all the elements." Stars make most elements heavier than helium, but the hydrogen in your body dates from the Big Bang, and lithium, beryllium and boron owe a lot to cosmic rays smashing into interstellar nuclei.

"Supernovae make all the gold." This was the textbook answer for decades. Modern models struggle to get ordinary core-collapse supernovae neutron-rich enough for a full r-process, and the 2017 kilonova gave direct evidence that mergers do the job. The share from other rare explosions is still debated.

"Iron is the most stable nucleus." Iron-56 has the lowest mass per nucleon, but nickel-62 has slightly higher binding energy per nucleon. The practical point for stars, that fusion past the iron peak costs energy, stays the same.

"The Big Bang made helium and nothing else." It also made deuterium, helium-3 and a small amount of lithium-7, and the measured abundances of these light nuclei are among the strongest tests of the Big Bang model.

"Nucleosynthesis is finished." It happens today in every star burning hydrogen, and every year somewhere in the observable universe neutron stars merge and massive stars explode. The universe's stock of heavy elements is still rising, which you can see in the list of the most abundant elements in the universe, where hydrogen and helium still dominate by a wide margin.

FAQ

Q: What is nucleosynthesis in simple terms? A: It is the process of building atomic nuclei from protons, neutrons and lighter nuclei. It happened in the first minutes after the Big Bang, and it still happens inside stars, in supernovae, in neutron star mergers and when cosmic rays strike interstellar gas.

Q: Which elements were made in the Big Bang? A: Hydrogen, deuterium, helium-3, helium-4 and a trace of lithium-7. By mass, the result was about 75% hydrogen and 25% helium. Heavier elements had to wait for stars.

Q: Why can't stars fuse elements heavier than iron for energy? A: Nuclei around iron and nickel are the most tightly bound. Fusing them into heavier nuclei absorbs energy instead of releasing it, so a massive star whose core turns into iron-peak material loses its energy source and collapses.

Q: Where does gold come from? A: Mostly from the rapid neutron-capture process. The 2017 neutron-star merger GW170817 provided direct evidence that such mergers make r-process elements, and strontium was identified in its spectrum in 2019. Whether other rare explosions also contribute significantly is still debated.

Q: What is the difference between the s-process and the r-process? A: In the slow process, nuclei capture neutrons one at a time over thousands of years, with time to decay between captures; it runs mainly in ageing AGB stars. In the rapid process, nuclei capture neutrons within seconds, far faster than they can decay, in extremely neutron-rich events such as mergers.

Q: Are we really made of star stuff? A: Yes. Apart from hydrogen, nearly every atom in your body, including the carbon, oxygen, calcium and iron, was made inside stars or in their explosions before the solar system formed about 4.6 billion years ago.

Sources

• Burbidge, Burbidge, Fowler & Hoyle, Synthesis of the Elements in Stars, Reviews of Modern Physics 29, 547 (1957)
• Kasen et al., Origin of the heavy elements in binary neutron-star mergers from a gravitational-wave event, Nature (2017)
• Watson et al., Identification of strontium in the merger of two neutron stars, Nature 574, 497–500 (2019)
• Bartos & Márka, A nearby neutron-star merger explains the actinide abundances in the early Solar System, Nature 569, 85–88 (2019)

Related Reading

• What Happens When a Star Dies
• What Is a Supernova and Why It Matters
• Stellar Evolution: The Complete Life Cycle of Stars
• Neutron Stars: The Most Extreme Objects in the Universe
• Gravitational Waves Explained