How Science Works | How Stars Build Chemical Elements

An element is identified by its proton number. Isotopes of that element contain different numbers of neutrons. Nucleosynthesis concerns the production of nuclei, so isotope identity matters even when ordinary chemical language simply says “carbon” or “iron”.

The early universe produced mainly hydrogen and helium, with small amounts of other light nuclei. Later stars inherited that material and processed part of it. Subsequent generations also inherited products from earlier stellar populations. A star is therefore both a consumer of an existing chemical inventory and a potential source of new material.

Most hydrogen nuclei in everyday matter do not require an earlier star to have manufactured them. Meanwhile, an individual heavy element can have several contributing routes. Different isotopes of the same element may have different production histories, and the relative contributions can vary with place and time.

Even knowing a likely route does not give an atom a recoverable address. Gas mixes, stars exchange material, ejecta travel and new systems form from blended reservoirs. The statement that an element has a stellar origin is usually a population-level inference, not a unique reconstruction of the particular star responsible for every atom in an object.

A good diagram starts with reservoirs rather than one dramatic explosion: early gas, ordinary stars, evolved-star winds, stellar explosions, compact-object events and interstellar gas. Arrows should show both processing and mixing. This makes the recycling visible from the beginning.

Contents · Next chapter

Inside suitable stars, nuclear reaction chains convert hydrogen into helium. The proton–proton chains and the CNO cycles take different paths; in the latter, carbon, nitrogen and oxygen nuclei participate in cycles whose net effect is hydrogen conversion. The relative importance of these routes depends on stellar conditions.

“Burning” is astronomical shorthand for nuclear processing. It is not combustion with oxygen. Nor do four protons simply undergo one ordinary four-particle collision to make helium. Intermediate reactions, weak interactions, emitted particles and reaction probabilities determine the actual pathway and rate.

The total energy budget can be illustrated with atomic masses. Assign the hydrogen-1 atomic mass 1.007825 u and helium-4 atomic mass 4.002603 u. Four hydrogen atoms minus one helium atom give a mass difference of 0.028697 u. Multiplying by approximately 931.5 MeV per atomic mass unit gives about 26.7 MeV.

This atomic-mass bookkeeping includes the net energy available after accounting consistently for electrons and positron annihilation. Some energy leaves in neutrinos, so the entire 26.7 MeV should not be assigned directly to local heating or eventual electromagnetic radiation. The division depends on the reaction chain.

A favourable energy difference also does not tell us how quickly the reactions proceed. Temperature, density, particle distributions and nuclear cross-sections matter. The slow parts of the reaction network help determine how a star can shine for a long time instead of consuming its accessible fuel instantaneously.

Contents · Previous chapter · Next chapter

After a star changes structure enough to ignite helium, the triple-alpha process can build carbon-12 from three helium-4 nuclei. The route passes through an unstable beryllium-8 intermediate and a resonant state in carbon. These details matter because the intermediate would otherwise usually disappear before the next step occurred.

Carbon-12 can then capture another helium nucleus to form oxygen-16. The competition between production and further processing helps set the carbon-to-oxygen mixture left by helium burning. It depends on reaction rates and on how material and energy move through the star.

Not every star reaches every later burning stage. Initial mass, mass loss, composition and interactions with a companion affect the route. Many stars end with carbon–oxygen-rich remnants rather than running through a complete sequence of advanced fuel stages.

For a bookkeeping exercise, begin with 120 helium-4 nuclei. If 30 carbon-12 nuclei are assembled, they consume 90 helium nuclei and leave 30. If 10 of those carbon nuclei subsequently capture helium, the final inventory is 20 carbon-12 nuclei, 10 oxygen-16 nuclei and 20 helium-4 nuclei.

The mass-number total remains 480: 20 × 12 + 10 × 16 + 20 × 4. The proton-number total also remains 240. These integer totals track nucleons and charge, not exact rest mass. Binding-energy changes mean exact masses are not obtained simply by adding mass numbers. The example is a conservation exercise, not a reaction-rate prediction or a literal description of a whole stellar core.

Contents · Previous chapter · Next chapter

Massive stars can undergo further stages involving carbon, neon, oxygen and silicon burning. “Silicon burning” is a network of rearrangements, captures and photodisintegrations, not simply two silicon nuclei repeatedly sticking together. At sufficiently high temperatures, many linked reactions compete.

The binding energy per nucleon reaches a broad maximum in the iron–nickel region. Rearranging light nuclei toward more tightly bound products can release energy, but fusion of iron-group material is not a generic new energy supply that can support an increasingly massive core indefinitely.

Production is only half the story. Material made deep inside a star may remain locked in a remnant or fall back after an explosion. Winds, envelope loss and explosive ejecta can return different parts of the star to surrounding gas. An element’s presence in an interior calculation therefore does not automatically equal its contribution to the next generation of stars.

Explosive burning can produce nuclei under rapidly changing conditions. Radioactive products can later decay into other elements; for example, nickel-56 decays through cobalt-56 to iron-56. Thus some iron observed much later was not ejected initially as stable iron.

The classic 1957 work by Burbidge, Burbidge, Fowler and Hoyle organised several distinct nuclear processes into an explanation of element abundances. Its enduring contribution is the need for multiple routes tied to nuclear physics and astrophysical environments. Modern work has revised sites and yields; the historical framework is not a licence to treat every original site assignment as settled today. Read the original synthesis.

Contents · Previous chapter · Next chapter

A neutron has no electric charge, so it does not face the same electrostatic repulsion as an incoming positively charged nucleus. A nucleus can capture a neutron and increase its mass number. Some resulting isotopes then undergo beta decay, converting a neutron into a proton while emitting an electron and an antineutrino.

In beta-minus decay, mass number stays the same while proton number increases by one. This combination of neutron captures and changes in proton number allows networks to move toward heavier elements without demanding that each step be an energy-releasing fusion reaction between two charged heavy nuclei.

The slow neutron-capture process, or s process, operates where captures are generally slow compared with relevant beta decays at unstable nuclei. Its path therefore tends to stay near the stable isotopes, although branching points complicate the simple picture. Evolved stars, including asymptotic giant branch stars, are important sites; other stellar environments also contribute.

The rapid process, or r process, occurs when captures can outpace many beta decays. The path can reach much more neutron-rich nuclei before subsequent decays move the products back toward stability. Neutron availability, temperature, expansion and nuclear properties all affect the final pattern.

“Slow” and “rapid” describe competition among rates, not a universal stopwatch dividing all environments into two classes. Some nuclei capture easily, others slowly; some decay quickly, others survive much longer. A reaction network follows these differences rather than assigning one identical behaviour to every isotope.

Contents · Previous chapter · Next chapter

Invent an unstable nucleus with a beta-decay half-life of 10 seconds. Its decay rate is ln2/10 ≈ 0.0693 per second. Suppose its neutron-capture rate is a constant 0.010 per second, and these are the only two competing exits from the state.

Under independent exponential waiting-time assumptions, the probability that capture happens before decay is λ_capture/(λ_capture + λ_decay). Here it is 0.010/(0.010 + 0.0693) ≈ 0.126, or 12.6%. Most nuclei in this invented ensemble would decay first.

Increase the capture rate to 1.0 per second while leaving the decay rate unchanged. The capture-first probability becomes 1/(1 + 0.0693) ≈ 0.935, or 93.5%. The same starting nucleus now overwhelmingly takes the capture route first.

This model concerns the next event, not the final abundance of an element after an entire stellar episode. It assumes constant rates and enough time for an exit. Real environments can cool, expand and exhaust their neutrons; photodisintegration and other reactions can open additional paths.

For a paper diagram, draw the starting isotope with two outgoing arrows. Label each arrow with its rate, then put the two calculated probabilities beside the corresponding cases. Do not label the probabilities “percentage of all cosmic heavy elements”. The diagram represents one branch in one hypothetical network, not a Galactic production census.

The exercise explains why nuclear measurements and astrophysical conditions have to be studied together. A laboratory decay property alone does not determine which branch wins in a particular neutron environment.

Contents · Previous chapter · Next chapter

A stellar spectrum can reveal nuclei that cannot plausibly have survived unchanged since the star’s birth. Technetium is a useful example because it has no stable isotope. Its identification in certain evolved stars supports relatively recent nuclear production and transport into observable layers, rather than an entirely inherited ancient surface composition.

A 1987 spectroscopic study by Little and colleagues examined technetium across evolved-star classes and connected its occurrence with stellar evolution. Such evidence links a nuclear clock with a stellar surface measurement. It still requires careful line identification and modelling; an absorption line is not a direct camera view of the production layer. Read the original observational study.

Compact-object events provide a different route. The light following the neutron-star merger GW170817 supplied spectra of the transient AT2017gfo. Watson and colleagues’ 2019 reanalysis identified strontium, a neutron-capture element, in those spectra. This connected heavy-element production with a specific merger environment rather than relying only on theoretical plausibility. Read the primary strontium study.

That identification does not mean every predicted element was individually detected. Spectral lines can blend in rapidly expanding ejecta, and atomic data and radiative-transfer models affect the interpretation. Broad light-curve evidence, one identified species and a complete isotopic inventory are different levels of information.

The strongest argument connects several observables: event identity, spectral evolution, temperature and luminosity changes, and predictions from models of expanding radioactive ejecta. Agreement across these features is more informative than a colourful artist’s image of newly made gold.

Contents · Previous chapter · Next chapter

Neutron-star mergers are important r-process sites, but their contribution through cosmic history must be compared with event rates, yields and the timing of enrichment. Other rare explosive environments remain relevant research subjects. An evidence-based account should not assign a fixed exclusive birthplace to every heavy nucleus.

Patel and colleagues’ 2025 study interpreted delayed gamma-ray emission after the 2004 giant flare of magnetar SGR 1806-20 using radioactive decay of newly synthesised r-process material. The predicted evolving emission matched the delayed signal, supporting a production route involving ejected neutron-star crust. The inference depends on the ejecta and nuclear-emission modelling; it is not a resolved measurement of every isotope or a complete census of the Galaxy’s heavy-element sources. Read the primary analysis.

This kind of result changes the question from “can one site make the elements?” to “how much does each viable site contribute, and when?” A rare source with a large yield can compete with a frequent source with a small yield. Delay times and where the ejecta are deposited also matter.

Production routes can even run in the opposite direction to assembly. Cosmic-ray collisions can fragment heavier nuclei, contributing importantly to lithium, beryllium and boron. Lithium also has primordial and stellar contributions. Therefore, an origin diagram that allows only upward steps toward larger nuclei misses part of the story.

The proper inventory is a network: build, decay, fragment, eject, mix and lock away. Each operation leaves a different sort of observable trace, and each must be included only where its conditions and evidence support it.

Contents · Previous chapter · Next chapter

Consider two imaginary source populations producing the same specified group of heavy nuclei. Source A occurs once per 10,000 years and ejects 0.020 solar masses of that group per event. Source B occurs once per 100 years and ejects 0.000010 solar masses per event.

Their average production rates are 0.0001 × 0.020 = 0.000002 solar masses per year for A, and 0.010 × 0.000010 = 0.0000001 for B. Despite being much rarer, A supplies twenty times as much in this constructed comparison. Its fraction of the combined production is 20/21 ≈ 95.2%.

These values are deliberately hypothetical. They must not be mistaken for measured merger or magnetar rates. The point is that a spectacular event, a high rate or a large yield cannot independently determine a population’s contribution. The product matters, and its uncertainties matter too.

Even that product is incomplete for chemical evolution. Suppose all B ejecta are retained in a gas reservoir but only 10% of A ejecta remain there. The retained rates become 0.0000002 and 0.0000001 solar masses per year, so A’s share drops to two-thirds. Transport and retention can change the composition of the next generation without changing the underlying nuclear yield.

Real modelling also follows finite stellar lifetimes, changing rates, gas inflow and outflow, radioactive decay and spatially incomplete mixing. A measured abundance pattern is the accumulated result of that history. Recovering it is an inverse problem, often with more than one plausible solution.

Contents · Previous chapter · Next chapter

Start by identifying the element and isotope where possible. Ask which reactions can produce it, which environment supplies the conditions, what ejects it and what observation tests that connection. Finally distinguish a source’s demonstrated ability to produce material from its inferred share of the total cosmic inventory.

Does hydrogen burning mean chemical fire? No; it is nuclear processing. Did our helium-to-carbon exercise conserve exact rest mass? It conserved nucleon number and charge; nuclear binding changes the rest-mass budget. What distinguishes slow and rapid neutron capture? The capture rates relative to competing decays and other network processes.

Does detecting strontium in a merger prove every gold atom came from a merger? No. Why can a rare source dominate a yield budget? A sufficiently large yield per event can outweigh a smaller event rate. Why might the next star inherit a different ratio from the gross production ratio? Ejecta retention, transport, mixing and timing alter what reaches its birth material.

The numerical reaction, branch and population examples here are original paper-only exercises. They are not laboratory instructions and require no radioactive material, radiation source or nuclear experiment. Their purpose is to keep conservation, rates and inference separate.

Return to Astrophysics, explore chemical organisation through the periodic-table lesson, or choose another mechanism in the How Science Works master guide. The elements become a readable history when their production routes and the evidence for those routes are kept together.

Contents · Previous chapter · Continue to Astrophysics