Why is there more matter than antimatter?

The leftover can be measured, because the annihilation products are still here as the microwave background. Comparing the number of ordinary particles in the universe to the number of photons gives about six in ten billion.

So matter outnumbered antimatter by something like one part in a billion. The pairs became light. The remainder became stars, planets and the people counting.

A tidy solution would be that the antimatter is elsewhere, in regions of its own. That was tested rather than assumed.

Wherever a matter region met an antimatter region, annihilation at the boundary would produce a gamma-ray glow. A 1998 analysis found that the expected signal exceeds what is observed unless our own matter domain is effectively the entire visible universe, which is to say that a patchwork is empirically dead anywhere we can see.

In 1967 Andrei Sakharov set out what any explanation must include. Some process must change the number of baryons. The laws must treat matter and antimatter differently. And the universe must fall out of thermal equilibrium while this happens, or any surplus produced is undone as fast as it is made.

In 1964 an experiment on neutral kaons found decays that should not happen if matter and antimatter behaved identically. Cronin and Fitch, two of its four authors, were given the 1980 Nobel Prize in Physics for the discovery.

The same violation has since been measured in B mesons, in 2019 in the decays of charm particles, and in 2025 in a baryon. In kaons and B mesons it fits the Standard Model’s account closely; whether the charm result does is still being settled.

It also falls short by roughly ten orders of magnitude. Canetti, Drewes and Shaposhnikov stop short of calling that a proof of impossibility, and add that the third condition fails as well: at the measured Higgs mass the electroweak transition is not first order, so the Standard Model never departs far enough from equilibrium.

A large class of proposals puts it in the neutrino sector. If neutrinos are their own antiparticles, heavy right-handed partners in the early universe could generate a lepton asymmetry, which sphaleron processes would then transfer into a baryon asymmetry.

Most baryogenesis models are hard to falsify, because an absent signal can be answered by raising the energy scale. Canetti, Drewes and Shaposhnikov count three that are testable in the strict sense that every parameter could be measured, and two of those are in the neutrino sector.

The Majorana neutrino those proposals need would show in a rare nuclear decay, two neutrons becoming two protons and two electrons with no neutrinos emitted at all. Every search so far has come back empty. Finding it would show that lepton number is broken and that neutrinos are their own antiparticles. That is a precondition for leptogenesis, not a confirmation of it, and lepton number broken at collider energies would typically wash out whatever standard high-scale leptogenesis had made.

Some versions of the story happen at energies a collider could probe, and are being squeezed by measurements of the electron’s electric dipole moment. Others can be placed at energies high enough to escape every experiment for centuries, leaving at most indirect traces.

The squeeze is real. A heavy particle violating time reversal, of the kind that could explain the asymmetry, would tilt the electron slightly. No tilt has been found. The limit set in 2018, tightened by a factor of about 2.4 in 2023, pushes a broad class of candidates, if they violate time reversal maximally, above anything the Large Hadron Collider can reach directly.

Which kind of answer this is determines whether it is a question about experiments or a question about cosmology.

Even a successful mechanism has to land on roughly six parts in ten billion rather than one part in a thousand or one part in a trillion. The number is an input baryogenesis is asked to reproduce, not something the known laws predict.

It is also possible that the asymmetry was never produced at all but set as an initial condition. That only survives if the universe did not inflate, or not for long. A long enough period of accelerated expansion dilutes any pre-existing asymmetry until it is negligible, which puts baryogenesis at reheating or later.