CloudSignal AI

Blog

The First Three Minutes: Where the Lightest Elements Come From

By Ashwin Rajendraprasad Cosmos

Nearly every hydrogen and helium nucleus in existence dates from the universe’s first few minutes, when a brief episode of nuclear fusion set their proportions and then shut itself off almost as soon as it began. The recipe was worked out on paper in 1948, and modern measurements match it so closely that the light elements now stand beside the cosmic microwave background as one of the pillars of the Big Bang. The phrase “the first three minutes” belongs to Steven Weinberg, whose 1977 book of that title made this earliest chapter of cosmic history famous; the chapter itself belongs to nuclear physics.

A reactor with a deadline

Run the expansion backwards and the universe grows hotter and denser without limit, toward a beginning that raises questions of its own. Run it forward from roughly one second of age and the physics becomes almost conventional: a gas of protons, neutrons, electrons, neutrinos, and photons at about ten billion kelvin, expanding and cooling according to nuclear and particle physics we test in laboratories. At those temperatures no nucleus survives; the radiation is energetic enough to blast apart any bound pair the instant it forms.

Every heavier nucleus has to begin with deuterium, a proton and a neutron bound together, and deuterium is fragile. Until the temperature fell below roughly a billion kelvin, a few minutes in, photons destroyed each deuteron as fast as it was made: the deuterium bottleneck. All the while a clock was running, because a free neutron decays with a mean lifetime of about fifteen minutes. When the bottleneck finally broke, fusion went fast: deuterium into helium-3 and tritium, then into helium-4, sweeping up essentially every neutron still alive. Within roughly twenty minutes the expanding gas had become too cool and too thin to fuse at all, and the reactor switched off for good.

The prediction of April 1948

The framework was laid out by George Gamow and his doctoral student Ralph Alpher in a one-page letter, “The Origin of Chemical Elements,” published in Physical Review on April 1, 1948. Gamow, unable to resist, added Hans Bethe to the author list purely to complete a pun on the Greek alphabet (Bethe had no hand in the work), so the byline reads Alpher, Bethe, Gamow: alpha, beta, gamma. The substance came from Alpher’s dissertation research: nuclei built up by successive neutron capture in a hot, expanding gas.

The original ambition was to make the entire periodic table this way. Nature declined; no stable nucleus exists at mass 5 or mass 8, so the chain stalls just past helium, leaving traces of lithium and little else. What survived the failed ambition was a sharp, checkable claim about the composition of everything there is: a hot early universe should have left behind roughly three-quarters hydrogen and one-quarter helium-4 by mass, with trace amounts of deuterium and lithium.

A quarter of the cosmos from one ratio

Why one quarter? Because the helium yield follows almost entirely from the ratio of neutrons to protons at the moment fusion began. Weak interactions held the two species in balance until the universe was about a second old; the ratio then froze near one neutron per six protons and drifted to about one per seven as neutrons decayed. Since nearly every surviving neutron ends up inside helium-4,

Yp≈2 (n/p)1+n/p≈0.25for n/p≈17,Y_p \approx \frac{2\,(n/p)}{1+n/p} \approx 0.25 \qquad \text{for } n/p \approx \tfrac{1}{7},

where YpY_p is the fraction of ordinary matter’s mass in helium-4 and n/pn/p is the neutron-to-proton ratio, the one number out of particle physics that fixes a quarter of the visible universe.

The measurements have caught up to the arithmetic. The most precise determination of primordial helium to date, from the LBT Yp Project at the Large Binocular Telescope (Yeh, Olive, Fields, Aver and collaborators, 2026), gives Yp=0.2458±0.0013Y_p = 0.2458 \pm 0.0013. Deuterium is sharper still: it is read from pristine gas clouds backlit by distant quasars, and standard nucleosynthesis, fed the baryon density measured by ESA’s Planck satellite, predicts a deuterium-to-hydrogen ratio of (2.442±0.040)×10−5(2.442 \pm 0.040)\times 10^{-5} (Launders, Giovanetti and Liu, 2026). Observation and prediction agree to within a few percent, closely enough that the field’s live argument is whether the residual, under two standard deviations, means anything at all.

The lithium problem

One ingredient refuses to cooperate. The same calculation that gets helium and deuterium right predicts a primordial lithium-7 abundance of about 5×10−105\times10^{-10} relative to hydrogen. Yet the oldest, most metal-poor stars of the galactic halo show a strikingly uniform lithium abundance (the Spite plateau, identified by François and Monique Spite in 1982) of about 1.6×10−101.6\times10^{-10}: a factor of three too little, as a 2025 study in Astronomy & Astrophysics by Oswaldo Miranda quantifies it. Candidate explanations range from astrophysics (old stars may slowly destroy the lithium in their own surface layers through diffusion and mixing) to revised nuclear inputs to new physics in the first minutes, and none is yet accepted as settled. Hydrogen, helium, and deuterium still carry the overwhelming evidential weight and agree with prediction, so the framework stands; lithium is the one light element it cannot yet account for.

The long cooling and the afterglow

When fusion ended, the universe was three-quarters hydrogen and a quarter helium by mass, and still a glowing plasma. For the next 380,000 years light could not travel far; free electrons scattered photons the way fog scatters headlights. Only when the temperature fell to about 3,000 kelvin could electrons settle into atoms, an event cosmologists call recombination, at which point the fog lifted and the universe became transparent. The light released in that moment has been traveling ever since, stretched by expansion into microwaves at 2.725 kelvin above absolute zero: the cosmic microwave background.

Its existence was predicted in 1948 by Alpher and Robert Herman in Nature, with an estimated temperature of about 5 K. Seventeen years later Arno Penzias and Robert Wilson, chasing a stubborn hiss in a Bell Labs horn antenna, measured an excess temperature of about 3.5 K at 4,080 MHz; Robert Dicke’s Princeton group supplied the cosmological interpretation in a companion paper, and the discovery earned Penzias and Wilson a share of the 1978 Nobel Prize in Physics. NASA’s COBE later showed the background to be the most perfect blackbody spectrum ever measured, and WMAP and Planck mapped its part-in-100,000 ripples: the same maps that supply the baryon density behind the deuterium test, and that anchor the precision cosmology now being stress-tested by surveys such as DESI.

Everything heavier came later

Big Bang nucleosynthesis ends just past lithium; the rest of the periodic table is the work of stars. The blueprint is the 1957 Reviews of Modern Physics paper “Synthesis of the Elements in Stars” by Margaret Burbidge, Geoffrey Burbidge, William Fowler, and Fred Hoyle, known ever since as B²FH, which laid out how stellar cores fuse carbon through iron and how heavier elements arise in neutron-capture processes over stellar lives and deaths. The enrichment began with the earliest stellar generations, in galaxies JWST is now watching assemble, and every generation since has salted the gas further. The calcium in your bones and the iron in your blood were forged in stars that died before the Sun existed; the hydrogen in your water dates from the first three minutes.

Nearly eight decades after the one-page letter, the account rests on two things open to inspection: a window of minutes, closed for 13.8 billion years, whose output we can still weigh, and an afterglow we can still see.

Sources / further reading

Written by Ashwin Rajendraprasad for CloudSignal AI.