Rabbit Holes a commonplace book,❦kept in the open
Explainer 29 min Updated 29 July 2026 Living document

Is it true that all the elements in the human body come from stars?

Mostly, but not entirely — and the exception is the most abundant thing in you. Around three out of every five atoms in your body are hydrogen, and no star has ever made a hydrogen nucleus. Stars destroy hydrogen; they do not create it. Those atoms were made in the first microsecond after the Big Bang and have been passed along ever since. By count, you are mostly older than stars.

The claim, tested

The line is Carl Sagan’s, more or less, and it has been repeated so often that it has stopped being a claim and become a mood. We are made of star-stuff. It is the kind of sentence that feels true in the chest before the brain has checked it, which is exactly the condition under which people stop checking.

So let us check it. Not to be clever — the sentence is very nearly right, and what it points at is genuinely one of the best things we know. But the way it is almost right is more interesting than the way it is right.

Start with what is actually in you. Take a body of seventy kilograms and sort it by mass, and four elements account for more than ninety-six per cent of you1: oxygen at about 65%, carbon at 18.5%, hydrogen at 9.5%, nitrogen at 3.2%. Then calcium, phosphorus, and a long diminishing tail of potassium, sulphur, sodium, chlorine, magnesium, and traces measured in milligrams.

Now sort the same body a different way — not by how much each element weighs, but by how many atoms there are. The order inverts.

Drag the figure sideways to read it all
The same body, sorted twice. Watch hydrogen's ribbon: a thin band on the left, the dominant block on the right. Every other ribbon has to cross it.
Freitas 1999, Table 3-1; origins after Johnson 2019

Hydrogen is the lightest thing there is, so its modest 9.5% of your mass turns out to be about 61% of your atoms1. Oxygen falls to roughly a quarter, carbon to a tenth. Three out of every five atoms in you is a single proton with an electron for company.

That distinction is not pedantry. It is the whole question. Because those two sortings have different answers to where did this come from.

Why by-count and by-mass disagree so violently

An oxygen atom is sixteen times heavier than a hydrogen atom. So a kilogram of oxygen contains a sixteenth as many atoms as a kilogram of hydrogen. Sorting by mass tells you what you are made of in the sense a butcher means it; sorting by count tells you what you are made of in the sense a chemist means it — how many individual pieces of each kind are in the machine.

For the question where were these made, count is the honest measure. Every atom has exactly one origin story regardless of what it weighs.

Here is the part the slogan gets wrong. No star has ever made a hydrogen nucleus. A hydrogen nucleus is one proton, and stars do not build protons — they consume them, fusing them into helium and releasing the energy that makes starlight. Every second, the Sun turns roughly six hundred million tonnes of hydrogen into helium. Stars are, on the whole, in the business of destroying hydrogen.

So where did yours come from? It was made in the first few minutes after the Big Bang, before a single star existed, and essentially every hydrogen atom now in your body has been sitting around unchanged since then2. The water you drank this morning is two-thirds primordial. The hydrogen in your DNA is 13.8 billion years old and has been through exactly nothing.

The rest of you is a different story, and this is where the slogan earns itself. The oxygen you are mostly made of, by weight, was forged in the cores of massive stars and scattered when they died. The carbon in every protein came out of helium fusion in stellar interiors. The calcium in your bones, the iron in your blood, the phosphorus holding your genome together — all of it was assembled inside stars, and none of it existed in the early universe at all3.

Even , the four grams of iron in your haemoglobin, sits at a very particular place in this story: right at the top of the curve where fusion stops paying, which is the reason massive stars die the way they do. (Iron is nearly the most tightly bound nucleus there is. Chapter four is about the “nearly”.)

And then there is the question nobody can answer. Matter and antimatter should have been produced in equal amounts and annihilated completely, leaving a universe of light and nothing else. Instead, for reasons that are genuinely unresolved, roughly one particle in a billion survived. You are made of the residue. Every atom in your body — primordial hydrogen and stellar oxygen alike — exists because of an asymmetry we have measured precisely and cannot explain.

So: are we made of stardust? By mass, overwhelmingly yes. By count, most of you is older than any star that has ever burned. And all of you is the leftover of something we do not understand.

The rest of this document is the argument for each of those three sentences, in that order, going down.

The first three minutes

The universe made hydrogen the way it made everything else in its first few minutes: by cooling.

That sounds like an anticlimax, and it is worth sitting with why it is not. Every process below is a process of stopping. The early universe was hot enough that nothing could stay bound — build a nucleus and the next collision breaks it. The history of the first three minutes is the history of the temperature falling past a series of thresholds, and at each one, something that had been impossible becomes permanent.

Drag the figure sideways to read it all
Sixty-one powers of ten, drawn to scale. Everything that decides the composition of ordinary matter is over inside the first twenty minutes; the remaining thirteen billion years add the hairline at the top.
Cyburt et al. 2016; Planck 2018
01 / 06

Before 10 to the power -43 s, no theory we possess applies. Not “we have not worked it out yet” — the equations of general relativity and quantum mechanics give contradictory answers, and nobody knows which to believe. Nothing is drawn to the left of this line, because there is nothing honest to draw there.

02 / 06

By about 10 to the power -6 s, the universe has cooled enough for quarks to bind into protons and neutrons and stay bound. Every hydrogen nucleus in your body dates from this moment. A hydrogen nucleus is a single proton; it was made here, and nothing since has changed it.

03 / 06

At about one second, neutrinos stop interacting and fly free. They are still flying: around 300 of them sit in every cubic centimetre of space, including the space your body is currently occupying. At that same moment the ratio of neutrons to protons freezes at about one to six — and that ratio decides how much helium the universe will ever have.

04 / 06

Then nothing happens for three minutes. Protons and neutrons collide constantly and form deuterium, and every deuterium nucleus is smashed apart by a photon almost immediately. This is the Deuterium Deuterium is one of two stable isotopes of hydrogen; the other is protium, or hydrogen-1, 1H. The deuterium nucleus (deuteron) contains one proton and one neutron, whereas the far more common 1H has no neutrons. Wikipedia CC BY-SA · read 2026-07-29 , and it is the reason the universe did not simply fuse itself into iron in its first hour.

05 / 06

Below about a billion kelvin, deuterium survives — and the queue that has been building for three minutes empties at once. Almost every available neutron ends up locked inside . Twenty minutes later it is over: the universe is too cool and too thin to fuse anything else. The result is roughly 75% hydrogen and 25% helium by mass2, and it will stay that way for a very long time.

06 / 06

Now look at the whole width of it. The band settles inside the first twenty minutes and then holds, flat, for thirteen billion years — three-quarters of the way across the figure without a single change. Only at the far right, after the first stars, does a hairline of everything else appear along the top. It is drawn again beneath at ten times magnification, because at true scale it is two per cent and you would not otherwise see it start. That hairline is most of your body by weight, and almost none of it by count.

That the theory works at all is the remarkable part. Big Bang nucleosynthesis has essentially one free parameter — the density of ordinary matter — and from it predicts the primordial abundances of four different nuclei. Deuterium, helium-3 and helium-4 agree with observation to within a few per cent2. The same parameter, measured completely independently from the cosmic microwave background, gives the same answer4. Two entirely different kinds of measurement, separated by 380,000 years of cosmic history, agreeing on one number.

Except for lithium, which does not agree at all

The fourth prediction is lithium-7, and it is wrong. Theory predicts about three times more primordial lithium than we actually observe in the oldest stars, and after four decades the discrepancy has not been resolved.

Proposed explanations fall into three groups: the stars destroy their own lithium in ways we have not modelled properly; the nuclear reaction rates are off; or something happened in the early universe that is not in the standard picture. The first is the most popular and the least exciting. None is established.

It is worth naming plainly, because it is the honest shape of the situation: three predictions land beautifully and one is off by a factor of three, and the theory is still almost certainly right. Being right about most things is what most correct theories look like from the inside.

Why there is anything at all

Go back to that first microsecond, because something happened there that the timeline draws but does not explain.

When energy converts into matter it produces particles and antiparticles in equal numbers. That is not a detail of one theory; it is what every experiment ever run has shown. Build a proton in an accelerator and you build an antiproton alongside it. And when a particle meets its antiparticle, both vanish into photons.

So the early universe should have annihilated itself completely. Equal amounts, perfectly cancelling, leaving a cosmos of light and nothing else. No hydrogen. No stars. No oxygen, no carbon, nobody.

Instead, for about every billion particle–antiparticle pairs, one particle was left over. Everything you have ever seen or touched — the whole visible universe — is that residue4.

Drag the figure sideways to read it all
The void is drawn at true scale. It is the largest thing in the figure because it is the largest thing in the problem.
Planck 2018 for the measured ratio; Standard Model estimate after Sakharov 1967 and subsequent work

Andrei Sakharov worked out in 1967 what any explanation would have to provide5. Three conditions: some process must be able to change the number of baryons; the laws of physics must treat matter and antimatter differently; and the universe must be out of thermal equilibrium while it happens. It is a remarkable piece of reasoning — three and a half pages, written without knowing the mechanism, that still frames the entire field.

Two of the three we have. The universe expanding and cooling is never in perfect equilibrium, so the third comes free. And the Standard Model does contain processes that violate baryon number — Sphaleron A sphaleron is a static (time-independent) solution to the electroweak field equations of the Standard Model of particle physics, and is involved in certain hypothetical processes that violate baryon and lepton numbers. Wikipedia CC BY-SA · read 2026-07-29 .

The problem is the middle condition. Matter and antimatter are treated differently — CP violation was found in kaons in 1964 and has been measured many times since. But the amount of CP violation In particle physics, CP violation is a violation of CP-symmetry : the combination of C-symmetry and P-symmetry. CP-symmetry states that the laws of physics should be the same if a particle is interchanged with its antiparticle (C-symmetry) while its spatial coordinates are inverted. Wikipedia CC BY-SA · read 2026-07-29 the Standard Model contains is too small by nearly nine orders of magnitude. Not marginally short. Short by a factor approaching a billion.

What “we don’t know” means here, precisely

It does not mean there are no candidate explanations. There are many: leptogenesis, in which an asymmetry among neutrinos is converted into an asymmetry among quarks; electroweak baryogenesis with physics beyond the Standard Model; asymmetries seeded before or during inflation.

What it means is that none of them has been confirmed, several would require particles nobody has found, and the experiments that would distinguish between them — searches for neutrinoless double beta decay, for permanent electric dipole moments, for CP violation in neutrino oscillations — have not yet returned a verdict.

This is the healthy kind of open question: sharply posed, with a measurable answer, and with instruments currently pointed at it. It is not a mystery. It is a gap, and people are working in it.

This is worth sitting with, because it is the single most load-bearing unknown in this entire document. Everything that follows — every nucleus, every star, every element in your body — is downstream of one part in a billion surviving for reasons nobody can currently write down.

How a star burns

Here is a problem that should have stopped astrophysics in its tracks, and for a while did.

Two protons repel each other. To fuse, they must come close enough for the strong force to take over — around a femtometre — and to get that close they must overcome an electrostatic barrier of roughly a million electronvolts. The temperature at the centre of the Sun is about 15.7 million kelvin6, which sounds enormous and corresponds to a typical particle energy of about 1.4 keV.

That is not nearly enough. It is short by a factor of about seven hundred. The Sun, on a classical account, cannot be burning. And yet.

Drag the figure sideways to read it all
Neither curve alone permits fusion. Their overlap — narrow, and about a million times smaller than either — is the entire energy budget of the Sun.
Computed from the standard forms; solar core temperature from Bahcall et al. 2005
01 / 04

The falling curve is how many protons have a given energy. Most sit near the typical thermal energy; the number with much more falls away exponentially. Waiting for a proton energetic enough to climb the barrier by brute force means waiting far longer than the universe has existed.

02 / 04

The rising curve is the chance of getting through the barrier without climbing it. In 1928 George Gamow George Gamow was a Soviet and American polymath, theoretical physicist and cosmologist. He was an early advocate and developer of Georges Lemaître's Big Bang theory. Wikipedia CC BY-SA · read 2026-07-29 showed that a quantum particle has a small but non-zero probability of appearing on the far side of a barrier it could never surmount7. He was explaining alpha decay. Within a year, others realised it ran in reverse — and that it was why stars shine.

03 / 04

Fusion needs both: a proton energetic enough to be worth counting, and lucky enough to tunnel. Multiply the two and you get a narrow window — the Gamow factor The Gamow factor, Sommerfeld factor or Gamow–Sommerfeld factor, named after physicists George Gamow and Arnold Sommerfeld, is a probability factor for two nuclear particles' chance of overcoming the Coulomb barrier in order to undergo nuclear reactions, for example in nuclear fusion. Wikipedia CC BY-SA · read 2026-07-29 — sitting at about 6 keV. Below it there is no tunnelling; above it there are no protons. The Sun burns in the gap.

04 / 04

Now watch what temperature does. Push the core hotter and the peak moves right and grows explosively — the proton–proton rate scales as roughly the fourth power of temperature, and the CNO cycle that dominates in heavier stars scales as the seventeenth. This is why stars are stable: a small expansion cools the core and throttles the reaction, hard. The Sun is a thermostat.

The size of that peak is worth dwelling on. It is around a millionth of either curve at its own maximum — which is to say the Sun runs on a rounding error, and it is the reason it has lasted long enough for anyone to notice. A star that burned its hydrogen efficiently would be a brief and useless thing.

The forge

Fusing hydrogen into helium releases energy. So does fusing helium into carbon, and carbon into oxygen, and on up. If that continued indefinitely a star would burn forever. It does not, and the reason is a single curve.

Drag the figure sideways to read it all
Measured, not modelled. The spike at helium-4 and the steps at carbon and oxygen are real features of the nuclear landscape, and a smooth fitted curve would erase the most interesting thing on the plot.
Measured values, AME2020 (Wang et al. 2021)

Binding energy per nucleon is how much energy you would have to supply to pull a nucleus apart, divided by the number of pieces. Higher means more tightly held. The curve rises steeply from hydrogen, climbs through helium and carbon and oxygen, flattens across the iron group, and then declines all the way to uranium8.

Everything about stellar death follows from that shape. Below the peak, combining nuclei moves you up the curve and releases the difference — fusion pays. Above the peak it moves you down, and fusion costs. A massive star burning through successive fuels is climbing toward the maximum, and when its core reaches iron there is nothing left that releases energy by fusing.

The core has been holding the star up by radiating. It stops. And the collapse that follows takes less than a second.

Iron-56 is not actually the most bound nucleus

It is the one everyone names, and it is very nearly right, which is the most dangerous kind of wrong.

By binding energy per nucleon the order at the top is at 8.7945 MeV, then at 8.7922, then at 8.79038. Nickel-62 wins, by about two parts in ten thousand.

Two separate things are being confused when iron-56 is called “the most stable nucleus”, and both are worth separating.

The first is which measure you use. Iron-56 does win on a different one: it has the lowest mass per nucleon of any nuclide. Nickel-62 is more tightly bound per nucleon, but a nickel nucleus contains more neutrons, and neutrons are heavier than protons. Binding energy per nucleon and mass per nucleon are not the same ranking, and iron-56 tops the second.

The second is why the universe is full of iron-56 rather than either. That is not about stability at all — it is about the route. Silicon burning in a collapsing core proceeds through reactions that favour , which then decays to and on to . Abundance is set by the path, not the destination. And that decay chain has a visible consequence: it is what powers the fading light curve of a type Ia supernova.

The distinction changes nothing about why fusion stops. It is here because a document that repeats a convenient near-truth in the one place it is easy to check has not earned trust anywhere else.

Making the heavy things

So fusion stops at iron. But there is gold in the ground, iodine in your thyroid, uranium in the rocks under your feet. Something built them, and it was not fusion.

The answer is neutrons. A neutron has no charge, so it feels no Coulomb barrier — it can walk into a nucleus at any energy at all. Capture enough of them and the nucleus becomes unstable, a neutron converts into a proton, and you have climbed one step up the periodic table. Do it repeatedly and you can build anything.

There are two ways to do it, and they differ only in speed.

Drag the figure sideways to read it all
Two routes to the same place. The slow one never leaves the valley; the fast one leaves it entirely and falls back.
Line of beta stability from the standard semi-empirical relation; process paths schematic after B²FH 1957

The S-process The slow neutron-capture process, or s-process, is a series of reactions in nuclear astrophysics that occur in stars, particularly asymptotic giant branch stars. The s-process is responsible for the creation (nucleosynthesis) of approximately half the atomic nuclei heavier than iron. Wikipedia CC BY-SA · read 2026-07-29 — s for slow — happens inside ageing low-mass stars, where neutrons arrive perhaps once every few thousand years. That is slow enough that any unstable nucleus decays before the next neutron shows up, so the path never strays from stability. It builds patiently up to bismuth and stops.

The R-process In nuclear astrophysics, the rapid neutron-capture process, also known as the r-process, is a set of nuclear reactions that is responsible for the creation of approximately half of the atomic nuclei heavier than iron, the "heavy elements", with the other half produced largely by the s-process. Wikipedia CC BY-SA · read 2026-07-29 — rapid — is the opposite. Neutrons arrive so fast, in something so violent, that nuclei are driven far into neutron-rich territory before they have any chance to decay. When the flux stops, the whole overloaded population decays back toward stability at once. This is the only way to make anything above bismuth, which means every atom of thorium and uranium in the Earth was made this way.

The obvious question is where. And the honest answer is that until 2017, nobody had ever seen it happen.

What B²FH got right in 1957

The framework above is not new. In 1957 Margaret Burbidge, Geoffrey Burbidge, William Fowler and Fred Hoyle published B2FH paper The B2FH paper was a landmark scientific paper on the origin of the chemical elements. The paper's title is "Synthesis of the Elements in Stars", but it became known as B2FH from the initials of its authors: Margaret Burbidge, Geoffrey Burbidge, William A. Fowler, and Fred Hoyle. Wikipedia CC BY-SA · read 2026-07-29 laying out essentially every process by which stars build elements — hydrogen burning, helium burning, the alpha process, e-process, s-process, r-process, p-process9. It is known by their initials, B²FH, and it is one of the most complete papers anyone has written in any field.

What they could not do was say where the r-process happens. They knew it needed an enormous neutron flux in a very short time; they suggested supernovae. It took sixty years to get a direct look, and the answer turned out to include something they could not have anticipated, because neutron-star mergers had not been conceived of.

Hoyle deserves a second note. He was the one who insisted that carbon must have a specific excited state, because otherwise stars could not make carbon and he was demonstrably made of it. He was right, the state was found, and the argument remains one of the very few successful predictions ever made from the fact of one’s own existence. He also coined the phrase “Big Bang”, on the radio, for a theory he opposed until he died.

On 17 August 2017, LIGO and Virgo detected two neutron stars spiralling into each other 130 million light years away. Within twelve hours, telescopes across the world had found the afterglow10. The Kilonova A kilonova is a transient astronomical event that occurs in a compact binary system when two neutron stars (BNS) or a neutron star and a black hole collide. Wikipedia CC BY-SA · read 2026-07-29 faded in a way that matched the radioactive decay of freshly made heavy elements. Modelling the light curve gives an ejected mass of a few hundredths of a solar mass of material heavier than iron11 — thousands of Earth-masses of it, from a single collision. Estimates differ between analyses, which is what you would expect from one event modelled several ways.

That is a direct observation of the r-process running. It is also one event, and one event cannot establish a rate. Whether mergers account for all of the r-process, most of it, or only part — with rare magnetically-driven supernovae making up the rest — is genuinely unsettled. The reason it matters is timing: mergers take a long time to happen after stars form, and there is r-process material in some very old stars that seems to have arrived too early.

The census

Put it all together and the periodic table stops being a chart of chemical properties. It becomes a map of production histories — at least six of them, overlapping.

Drag the figure sideways to read it all
Two things at once: where each element comes from, and how well that is known. The hatching is not decoration — it is most of the table.
Johnson 2019

The first thing worth noticing is how much of the table is hatched. The broad picture is secure: light elements from the Big Bang, the middle from stars, the heaviest from catastrophes. But element by element, the split between competing sources is frequently uncertain, and most uncertain exactly where the s- and r-processes overlap.

The second thing is lithium, drawn voided. It is the only element on the table whose primordial abundance we predict confidently and measure confidently, and get two different answers.

Now find the outlined cells — the elements in you. They cluster in the well-understood part of the table, which is a piece of luck rather than a fact about biology: life is built from common elements, and common elements are the ones we have had the most opportunity to study.

How we know

At this point a reasonable reader should be getting suspicious. Nobody watched a carbon atom being made. Nobody has been inside a star. These are claims about events that happened billions of years ago, at temperatures no instrument could survive, and they are being stated with what looks like confidence.

So here is the machinery. Every claim in this document is the end of a chain, and the chains are short enough to check.

Each chain runs from something someone actually measured to a claim made above. The dashed rail marks the one that rests on a single event.
See individual chain notes

The first of those chains has a story attached that is worth telling properly.

In 1925 Cecilia Payne-Gaposchkin Cecilia Payne-Gaposchkin was a British-born American astronomer and astrophysicist. Her work on the cosmic makeup of the universe and the nature of variable stars was foundational to modern astrophysics. Wikipedia CC BY-SA · read 2026-07-29 submitted a doctoral thesis at Radcliffe applying new quantum theory to stellar spectra. Her conclusion was that stars are overwhelmingly hydrogen and helium — that the Sun is about a million times richer in hydrogen than in the metals that dominate the Earth12.

This contradicted the settled view that stars had roughly the composition of the Earth’s crust. Henry Norris Russell, the most senior astronomer in America, advised her the result could not be right. She added a sentence to her own thesis calling the hydrogen and helium abundances “almost certainly not real”.

They were real. Russell reached the same conclusion four years later by a different route, credited her, and is nonetheless the name usually attached to the discovery. Otto Struve later called her thesis the most brilliant ever written in astronomy.

That is the foundation of everything in the chapters above. If stars were not mostly hydrogen, none of the rest of this works.

The edge

Here is where we started.

Drag the figure sideways to read it all
The same figure that opened this piece, carrying everything the reading added to it.
Freitas 1999, Table 3-1; origins after Johnson 2019

The answer to the question in the title is: mostly, with a large and specific exception, and resting on a foundation nobody can currently explain.

By mass, you are stellar. The oxygen that is most of your weight, the carbon in every protein, the calcium in your bones, the iron carrying oxygen through your blood — all of it was assembled inside stars and scattered when they died. That part of the slogan is true, and it is remarkable, and it deserves the awe it gets.

By count, you are primordial. Three out of five of your atoms are hydrogen, made in the first microsecond, older than every star that has ever existed. Those atoms have been through nothing. They are the oldest things you will ever touch, and you are made mostly of them.

And all of it is a residue. Every atom in your body, stellar and primordial alike, exists because roughly one particle in a billion failed to annihilate for reasons that remain unexplained.

Three questions this document reached and could not close:

  • Why is there any matter at all? Sakharov’s conditions are known; the Standard Model supplies too little CP violation by nearly nine orders of magnitude. Candidate explanations exist. None is confirmed.
  • Where does the rapid neutron-capture process mainly happen? Neutron-star mergers certainly do it. Whether they do all of it, and whether they happen early enough to explain heavy elements in the oldest stars, is open.
  • Why is there three times less lithium than there should be? The oldest stars contain about a third of the predicted primordial lithium. Forty years of proposed resolutions, none established.

None of these is a gap in the popular account that dissolves on closer reading. They are gaps in the actual science, and they sit inside your left hand along with everything else.

That seems a better ending than the slogan. You are not simply made of stardust. You are made of the first three minutes, and of stars, and of one improbable accounting error — and the parts we cannot explain are not at the edges. They are in the hydrogen, which is most of you.

This document is alive

This is not finished, and is not meant to be. It changes when a measurement improves, when I get something wrong, or when I change my mind. Everything that has changed since it was first published is below.

29 Jul 2026

First published. Nothing has changed yet.

Sources

01
Nanomedicine, Volume I — Table 3-1: Elemental composition of the human body — Robert A. Freitas Jr., 1999

The standard tabulation of human elemental composition by mass and by atom count.

02
Big Bang Nucleosynthesis: Present Status — Cyburt, Fields, Olive & Yeh, 2016

Review of primordial abundances and how well theory and observation agree.

03
Populating the periodic table: Nucleosynthesis of the elements — Jennifer A. Johnson, 2019

Where each element is made, and how confidently that is known.

04
Planck 2018 results. VI. Cosmological parameters — Planck Collaboration, 2020

The measured baryon density, and the age of the universe to better than a per cent.

05
Violation of CP invariance, C asymmetry, and baryon asymmetry of the universe — Andrei Sakharov, 1967

Three and a half pages setting out the three conditions any explanation of the matter surplus must satisfy.

06
New Solar Opacities, Abundances, Helioseismology, and Neutrino Fluxes — Bahcall, Serenelli & Basu, 2005

The standard solar model — source of the 15.7 million kelvin core temperature.

07
Zur Quantentheorie des Atomkernes — George Gamow, 1928

Tunnelling through the Coulomb barrier — written about alpha decay, and the reason we understand why stars shine.

08
The AME2020 atomic mass evaluation (II). Tables, graphs and references — Wang, Huang, Kondev, Audi & Naimi, 2021

The evaluated nuclear mass table. Every binding energy in this piece is measured, not modelled.

09
Synthesis of the Elements in Stars — Burbidge, Burbidge, Fowler & Hoyle, 1957

B²FH. A hundred and four pages that laid out nearly every process by which stars build elements, and largely got it right.

10
Multi-messenger Observations of a Binary Neutron Star Merger — LIGO, Virgo and 70 observatories, 2017

The single direct observation of rapid neutron capture happening. One event, watched by almost every telescope on Earth.

11
Origin of the heavy elements in binary neutron-star mergers from a gravitational-wave event — Kasen, Metzger, Barnes, Quataert & Ramirez-Ruiz, 2017

The kilonova light curve read as an element inventory. Modelled ejecta masses are of order a few hundredths of a solar mass; estimates vary between analyses.

12
Stellar Atmospheres: A Contribution to the Observational Study of High Temperature in the Reversing Layers of Stars — Cecilia H. Payne, 1925

The doctoral thesis that found stars are overwhelmingly hydrogen, and was pressured into calling the result "almost certainly not real".

Hover-glosses on technical terms are summaries from Wikipedia, used under CC BY-SA 4.0. Each card links to the article it came from and records the date it was read.