Leptogenesis: How Neutrinos May Explain Why the Universe Contains Matter

Every atom on Earth is made of matter. Every star observed through a telescope is made of matter. Every galaxy imaged by Hubble and JWST is made of matter. Yet the Standard Model, which describes the laws of physics governing all those objects, treats matter and antimatter as exactly symmetric. For every rule that creates a particle, there is an equally valid rule that creates an antiparticle. The early universe, according to the same Standard Model, should have produced matter and antimatter in equal amounts, annihilated them as it cooled, and left behind only photons.

The fact that we exist — that the universe contains atoms, planets, and people at all — is therefore a problem. The ratio of surviving matter to the radiation photons produced in that annihilation is tiny but non-zero: about one part in a billion. This baryon asymmetry is one of the best-measured quantities in cosmology and one of the least well-explained in particle physics.

Leptogenesis is the name of the most theoretically compelling scenario for how this asymmetry came to exist. It ties the matter-antimatter imbalance to CP-violating processes in the lepton sector — specifically, to the decays of very heavy right-handed neutrinos in the early universe. If leptogenesis is correct, the same physics that gives rise to neutrino mass also gives rise to the existence of atoms. This essay explains the idea, the three Sakharov conditions it satisfies, the observational constraints it faces, and how future neutrino experiments could indirectly support or exclude it.

The asymmetry to be explained

Precision cosmological observations, particularly from the Planck satellite’s measurements of the cosmic microwave background and the primordial abundances of light elements produced in Big Bang nucleosynthesis, converge on a single number: the baryon-to-photon ratio in the present universe is ${\eta }_B=\frac{n_B-n_{\bar{B}}}{n_{\gamma }}\approx 6.1\times 10^{-10}$ In the early universe, $n_{\bar{B}}$ was essentially equal to $n_B$ (both large). As the universe cooled below $\sim 10^{12}$ K, matter and antimatter annihilated on contact, leaving behind the photons of the CMB and a tiny residual of one kind. That residual became all the baryonic matter in the universe today.

The question is why the residual existed in the first place. The Standard Model is CPT-symmetric and has tight constraints on explicit symmetry breaking. Without some mechanism that dynamically generates a preference for matter over antimatter in the early universe, ${\eta }_B$ should be zero.

The Sakharov conditions

In 1967, Andrei Sakharov identified three necessary conditions for dynamically generating a baryon asymmetry from an initially symmetric universe:

1. Baryon-number violation — some process must change the net baryon number.
2. C and CP violation — the interaction must treat particles and antiparticles differently at the quantum level.
3. Departure from thermal equilibrium — if the system stays in equilibrium, every process and its CP-conjugate occur at equal rates, producing no net asymmetry.

The Standard Model technically satisfies all three conditions — but only nominally. Baryon-number violation occurs through electroweak sphalerons at temperatures above about 160 GeV. CP violation exists through the CKM quark mixing matrix. The universe departs from equilibrium during the electroweak phase transition as the Higgs field settles to its vacuum value.

The numerical problem is that the CKM CP violation is far too small. The asymmetry it could generate is roughly $10^{-20}$ — ten orders of magnitude below the observed $10^{-10}$. The electroweak phase transition, additionally, is not first-order in the Standard Model; it is a smooth crossover that provides only marginal departure from equilibrium.

Something beyond the Standard Model is therefore required to produce the observed asymmetry. The question is what.

Heavy right-handed neutrinos and the seesaw

The most economical extension of the Standard Model that explains both the observed neutrino masses and the matter-antimatter asymmetry adds three heavy right-handed neutrinos. Each of these right-handed states — commonly labelled $N_1$, $N_2$, $N_3$ — carries no Standard Model gauge charge and therefore does not participate in electromagnetic, weak, or strong interactions. They interact with the ordinary Standard Model fields only through a Yukawa coupling to the Higgs and the left-handed lepton doublet.

Because they carry no gauge charge, the right-handed neutrinos can have a Majorana mass term: a mass that identifies each right-handed state with its antiparticle. If this Majorana mass $M$ is very heavy — typically $10^9$ to $10^{14}$ GeV — the light observed neutrinos acquire masses of order $m_{\nu }\sim \frac{m_D^2}{M}$ where $m_D\sim 100$ GeV is the Dirac Yukawa coupling scale. This is the Type-I seesaw mechanism, and it naturally explains why observed neutrino masses are many orders of magnitude smaller than all other fermion masses.

The heavy right-handed neutrinos themselves cannot be produced at current accelerators and decay quickly, but they were abundantly produced in the hot early universe, when the temperature exceeded $M$. As the universe cooled, those $N_i$ decayed. Because Majorana particles are their own antiparticles, the decay products can include either a lepton (from $N\to {\ell }^{-}{H}^{+}$) or an antilepton (from $N\to {\ell }^{+}{H}^{-}$). The two channels are CP-conjugates of each other.

If the Yukawa couplings contain CP-violating phases — which they generically do — the two channels occur at slightly different rates. The asymmetry, per decay, is roughly $ϵ=\frac{\Gamma (N\to {\ell }^{-}{H}^{+})-\Gamma (N\to {\ell }^{+}{H}^{-})}{{\Gamma }_{\text{total}}}$ and is typically of order $10^{-6}$ in realistic models. Each $N$ decay therefore produces slightly more leptons than antileptons, accumulating a net lepton asymmetry in the universe.

The sphaleron conversion

The lepton asymmetry produced by heavy-neutrino decays is not directly the baryon asymmetry. They differ by the total lepton and baryon numbers carried by the relevant particles. The conversion happens through electroweak sphalerons — non-perturbative gauge-field configurations that can convert lepton number into baryon number above the electroweak phase transition.

Sphalerons are active in the temperature range $10^{12}\gtrsim T\gtrsim 10^{10}$ K (or roughly $100$ GeV to $10^9$ GeV in particle physics units). In this window, they efficiently translate lepton asymmetry into baryon asymmetry with a conversion factor of ${\eta }_B\sim \frac{28}{79}{\eta }_L\approx 0.35{\eta }_L$ for standard electroweak gauge content. Below the electroweak phase transition, sphalerons are exponentially suppressed and the conversion freezes out. The final baryon asymmetry is whatever the sphalerons had produced up to that point.

Thermal equilibrium and the wash-out problem

A subtlety that shapes the viable parameter space: while the heavy $N_i$ are decaying, the inverse process $\ell +H\to N$ is also occurring. If the interactions are rapid compared to the Hubble expansion rate, the population of $N_i$ stays in thermal equilibrium, and any asymmetry generated by decays is immediately washed out by the inverse processes.

For leptogenesis to succeed, the Hubble expansion must be fast enough — or equivalently, the Yukawa couplings weak enough — for the $N_i$ decays to be “out-of-equilibrium” in Sakharov’s sense. This condition translates into an upper bound on the Yukawa couplings and, through the seesaw relation, a lower bound on the heavy Majorana mass. The typical successful leptogenesis scenario has $M_1\gtrsim 10^9$ GeV.

Detailed numerical analyses of the Boltzmann equations governing the abundance of $N_i$ and the resulting lepton asymmetry show that the observed baryon asymmetry ${\eta }_B\approx 6\times 10^{-10}$ can be reproduced for CP asymmetries of order $10^{-6}$ and heavy-neutrino masses in the expected seesaw range — a remarkable consistency that is the principal reason leptogenesis is considered the default explanation of the matter-antimatter asymmetry in most modern extensions of the Standard Model.

The observational handle

The fundamental challenge of leptogenesis is that it operates at energies — $10^9$ to $10^{14}$ GeV — that are vastly inaccessible to any conceivable terrestrial accelerator. Direct confirmation is impossible. Indirect tests come from three sources:

Light neutrino masses. The seesaw relation $m_{\nu }\sim m_D^2/M$ ties the observed light masses to the heavy sector. Measuring the light masses precisely (KATRIN + cosmology + $0\nu \beta \beta$) constrains the product $m_D^2/M$ but not the two factors separately. Different high-scale scenarios producing the same light mass are observationally indistinguishable at low energies.

Neutrinoless double beta decay. Observing $0\nu \beta \beta$ would establish that neutrinos are Majorana particles, a necessary (though not sufficient) condition for the leptogenesis framework. A positive detection would not prove leptogenesis but would remove a key alternative — Dirac neutrinos, in which leptogenesis cannot proceed through this mechanism.

Low-energy leptonic CP violation. The Dirac CP phase ${\delta }_{CP}$ measured by DUNE and Hyper-Kamiokande is, in general, unrelated to the high-energy CP phases responsible for leptogenesis. In specific models (for example, flavor-symmetry-driven textures), the low-energy and high-energy CP violations are connected, and measurement of ${\delta }_{CP}$ provides model-dependent constraints. In the most generic seesaw scenarios, the low-energy phase is essentially uncorrelated with the high-energy phases that actually generate the asymmetry.

No single measurement, in other words, will decisively confirm or rule out leptogenesis. What it will do is accumulate consistency: if ${\delta }_{CP}$ is near $-\pi /2$, $0\nu \beta \beta$ is observed, and the neutrino mass is near cosmological bounds, the standard seesaw-plus-leptogenesis framework becomes increasingly preferred over alternatives.

Alternatives and variants

Several modifications and alternatives to the canonical Type-I leptogenesis scenario have been developed:

Resonant leptogenesis. If two or more heavy neutrinos are quasi-degenerate in mass — their splitting comparable to their decay width — the CP asymmetry can be resonantly enhanced to near unity, allowing leptogenesis at much lower scales (TeV range). This makes the heavy sector potentially accessible to future colliders.

Dirac leptogenesis. If neutrinos are Dirac rather than Majorana, a different mechanism using the chiral asymmetry between left- and right-handed neutrinos in the early universe can produce the required lepton excess. Fewer models, but not yet excluded.

Type-II / Type-III seesaw leptogenesis. Uses a triplet Higgs or fermion triplet rather than a singlet right-handed neutrino. Provides similar phenomenology but with different flavor patterns.

Baryogenesis through non-neutrino mechanisms. Electroweak baryogenesis (with extended Higgs sector), Affleck-Dine baryogenesis (using scalar field dynamics in supersymmetric theories), and grand-unified baryogenesis are all alive as alternatives. Most modern community opinion, however, places leptogenesis at the top of the list.

Why this matters

The matter-antimatter asymmetry is not a peripheral question in cosmology. It is the reason there are atoms to form galaxies, stars to produce the heavy elements we are made of, and planets to harbor life. If leptogenesis is the correct explanation, the existence of physical matter in the universe is a direct consequence of the same processes that give neutrinos mass — a remarkable unification of the extremely small (neutrino properties) and the extremely large (cosmic baryon asymmetry).

The experimental path to establishing leptogenesis is slow, indirect, and multi-decadal. But each step — a first-principles mass bound from KATRIN, a CP violation measurement from DUNE or Hyper-K, an eventual $0\nu \beta \beta$ detection or stringent null result — tightens the web of constraints around the remaining viable models. By the 2040s, a consensus may be possible. Until then, the most likely explanation for why the universe exists is that it came from neutrinos decaying asymmetrically when the universe was microseconds old.