Neutrino decoupling

Neutrino decoupling is the process in the early universe during which neutrinos cease to maintain thermal equilibrium with the surrounding plasma through weak interactions, transitioning to free-streaming particles that no longer scatter significantly off electrons, positrons, or other neutrinos.^[1] This decoupling occurs at a temperature of approximately 1–2 MeV, corresponding to about 1 second after the Big Bang and a redshift of roughly 10^9 to 10^10, when the expansion rate of the universe surpasses the weak interaction rates.^[2] Prior to decoupling, neutrinos are in thermal equilibrium via processes such as neutrino-electron scattering (ν e⁻ ↔ ν e⁻) and neutrino-antineutrino annihilation into electron-positron pairs (ν ν̄ ↔ e⁺ e⁻), but as the universe cools, these rates drop below the Hubble expansion rate, allowing neutrinos to decouple while still relativistic. Following decoupling, the subsequent annihilation of electron-positron pairs at around 0.5 MeV heats the photon bath but not the neutrinos, leading to a temperature ratio of T_γ / T_ν ≈ 1.4 today and establishing the cosmic neutrino background (CνB) with a present-day temperature of about 1.95 K.^[1] This asymmetry results in the effective number of relativistic neutrino species, N_eff, being precisely 3.044 in the Standard Model after accounting for finite-temperature quantum electrodynamic (QED) corrections and partial decoupling effects on low-momentum neutrinos.^[2] Electron neutrinos decouple slightly later than muon and tau neutrinos due to charged-current interactions, contributing an additional ~0.83% to the relativistic energy density compared to the instantaneous decoupling approximation. The implications of neutrino decoupling extend to key cosmological observables, influencing Big Bang nucleosynthesis (BBN) by setting the radiation energy density at T ~ 0.1 MeV, which affects light element abundances such as helium-4 (with a shift ΔY ≈ +10^{-4}) and deuterium. It also impacts the cosmic microwave background (CMB) power spectrum through N_eff contributions to the early universe's expansion history, with deviations from 3.044 probed by experiments like Planck and future CMB-S4.^[1] Furthermore, spectral distortions from incomplete thermalization—such as a ~1% enhancement in the neutrino number density—affect prospects for direct CνB detection via experiments like PTOLEMY, which aim to capture relic neutrinos on tritium targets.^[1] These effects underscore neutrino decoupling's role in bridging particle physics and cosmology, with ongoing refinements from quantum kinetic equation solutions improving predictions to uncertainties of 10^{-3}–10^{-4} in N_eff.^[2]

Background

Weak interactions in the early universe

In the early universe, weak interactions govern the behavior of neutrinos through processes mediated by the massive W and Z bosons of the electroweak sector of the Standard Model.^[1] Charged-current interactions, facilitated by the charged W bosons, involve flavor-changing processes such as the reaction \nu_e + n \leftrightarrow p + e^-, where an electron neutrino interacts with a neutron to produce a proton and electron, enabling conversions between neutrons and protons.^[1] Neutral-current interactions, mediated by the neutral Z boson, preserve flavor and include elastic scatterings like \nu + e \leftrightarrow \nu + e, allowing neutrinos to exchange energy and momentum with electrons without altering particle types.^[1] These processes ensure that neutrinos remain in close contact with the thermal plasma during the hot, dense phases of cosmic evolution. Neutrinos exist in three flavors—electron (\nu_e), muon (\nu_\mu), and tau (\nu_\tau)—and are treated as massless or nearly massless particles in the context of early universe epochs relevant to decoupling, consistent with the Standard Model where their masses are negligible compared to the temperatures involved.^[3] In weak interactions, only left-handed chiral neutrinos participate, reflecting the V-A (vector-axial vector) structure of the weak force, which couples exclusively to left-handed fermions and right-handed antifermions.^[1] This chirality selectivity is a fundamental feature of electroweak theory, distinguishing weak interactions from the parity-symmetric electromagnetic and strong forces.^[1] At temperatures T > 1 MeV, long after the electroweak symmetry breaking at around 100 GeV,^[4] the early universe is a relativistic plasma where weak interactions dominate the dynamics of leptons, outpacing electromagnetic interactions for neutrinos due to the latter's lack of electric charge and the short range of the weak force.^[3] In this regime, the density and energy scales ensure frequent weak scatterings, maintaining neutrinos in thermal equilibrium with electrons, positrons, and other particles until the expansion dilutes the interaction rates.^[5] The efficacy of these interactions is characterized by cross-sections that scale as \sigma \approx G_F^2 E^2, where G_F is the Fermi coupling constant (G_F \approx 1.166 \times 10^{-5} GeV^{-2}) and E is the typical center-of-mass energy of the colliding particles, typically on the order of the temperature T.^[1] This quadratic energy dependence arises from the point-like nature of weak interactions at low energies (below the electroweak scale), making the rates increase rapidly with temperature and ensuring robust coupling in the MeV regime.^[3]

Thermal equilibrium and interaction rates

In the relativistic plasma of the early universe, at temperatures exceeding approximately 1 MeV, neutrinos, photons, electrons, and positrons form a thermal bath where frequent particle interactions maintain a common temperature T for all species, ensuring thermal equilibrium through the rapid exchange of energy and momentum.^[1] This equilibrium distribution follows Fermi-Dirac or Bose-Einstein statistics, with neutrinos occupying a thermal spectrum characterized by zero chemical potential due to efficient weak processes that equalize particle and antiparticle abundances.^[1] The rate at which neutrinos interact with the plasma, maintaining this equilibrium, is quantified by the interaction rate \Gamma = n \langle \sigma v \rangle, where n is the number density of target particles (such as electrons or positrons, scaling as T^3), and \langle \sigma v \rangle is the thermally averaged product of the weak interaction cross-section \sigma and relative velocity v. For temperatures around 1 MeV, where the weak interactions are dominated by four-fermion processes mediated by the Fermi constant G_F, the cross-section scales as \sigma \sim G_F^2 T^2, leading to the approximate form \Gamma \approx G_F^2 T^5.^[1] To assess whether interactions suffice to preserve equilibrium against cosmic expansion, this rate is compared to the Hubble expansion parameter H, which governs the dilution of the plasma. In the radiation-dominated era, H \approx 1.66 \, g_*^{1/2} \, G_N^{1/2} \, T^2, where g_* counts the effective relativistic degrees of freedom (approximately 10.75 near 1 MeV, including contributions from photons, electron-positron pairs, and neutrinos), and G_N is Newton's constant (or equivalently, H \approx 1.66 \, g_*^{1/2} \, T^2 / M_\mathrm{Pl} with reduced Planck mass M_\mathrm{Pl}).^[1] As the universe expands and cools below about 1 MeV, the steeper temperature dependence of \Gamma (fifth power) compared to [H](/page/H+) (second power) causes the interaction rate to fall below the expansion rate, transitioning the regime from interaction-dominated (where \Gamma \gg [H](/page/H+)) to expansion-dominated (where \Gamma \ll [H](/page/H+)). This inefficiency in weak interactions initiates neutrino decoupling, freezing their distribution while the remaining plasma continues to thermalize.^[1]

Decoupling process

Derivation of decoupling temperature

The decoupling temperature T_\mathrm{dec} for neutrinos in the early universe is defined as the temperature at which the weak interaction rate \Gamma(T) equals the Hubble expansion rate H(T), marking the transition from thermal equilibrium to free-streaming behavior. This condition, \Gamma(T_\mathrm{dec}) = H(T_\mathrm{dec}), is solved using the Boltzmann equation for neutrino distribution functions, incorporating collision terms from weak processes such as neutrino-electron scattering (\nu e \leftrightarrow \nu e), annihilation (\nu \bar{\nu} \leftrightarrow e^+ e^-), and exchange. The interaction rate \Gamma(T) arises from the weak cross-sections, scaled by the number densities and velocities of particles in the plasma. In the relativistic limit at temperatures T \sim 1 MeV, where T \gg m_e (electron mass), the leading contribution is \Gamma(T) \approx \frac{9}{4} G_F^2 T^5 for neutral-current processes, with G_F \approx 1.166 \times 10^{-5} GeV^{-2} the Fermi coupling constant; charged-current contributions add similar terms for electron neutrinos. The Hubble rate in the radiation-dominated era follows from the Friedmann equation,

H(T) = \sqrt{\frac{8\pi G \rho}{3}} = \frac{1.66 \sqrt{g_*} \, T^2}{M_\mathrm{Pl}},

where M_\mathrm{Pl} \approx 1.22 \times 10^{19} GeV is the reduced Planck mass, G is Newton's constant, and g_* = 10.75 accounts for the effective relativistic degrees of freedom (2 for photons, $7/8 \times 4 for electrons and positrons, and $7/8 \times 12/2 = 10.5/2 wait, precisely 2 + (7/8)(4 + 6) = 10.75, including three neutrino flavors). To find T_\mathrm{dec}, set \Gamma(T) = H(T):

G_F^2 T^5 \approx \frac{\sqrt{g_*} \, T^2}{M_\mathrm{Pl}}.

Solving for T,

T_\mathrm{dec}^3 \approx \frac{\sqrt{g_*}}{G_F^2 M_\mathrm{Pl}},

which yields T_\mathrm{dec} \approx 1 MeV upon inserting numerical values (g_* = 10.75, G_F, M_\mathrm{Pl}). More precise numerical integration of the Boltzmann equation, including full Fermi-Dirac statistics for blocking factors and Pauli suppression, refines this to T_\mathrm{dec} \approx 1{-}2 MeV. The corresponding time of decoupling is obtained from the radiation-dominated relation t = \frac{1}{2H(T)}, giving t_\mathrm{dec} \approx [1](/page/1) s at T_\mathrm{dec} \sim [1](/page/1) MeV, consistent with entropy conservation and scale factor evolution a \propto t^{1/2}. Flavor-specific effects arise because electron neutrinos (\nu_e) experience additional charged-current interactions with electrons, positrons, protons, and neutrons, increasing their \Gamma(T) and delaying decoupling to T_{\nu_e} \approx 0.8 MeV, while muon and tau neutrinos decouple earlier via neutral currents alone. Finite-temperature quantum corrections, including anharmonic terms in the weak rates and electron mass dependence (m_e / T \approx 0.5 at decoupling), shift the neutral-current decoupling to T_\mathrm{dec} \approx 1.5 MeV; these effects enhance the rates by up to 4-5% through thermal field theory propagators and radiative corrections.

Post-decoupling neutrino evolution

Following neutrino decoupling at temperatures around 2 MeV, the neutrinos remain in thermal equilibrium among themselves through weak interactions but decouple from the electromagnetic plasma consisting of electrons, positrons, and photons.^[1] Their distribution function adopts a Fermi-Dirac form with zero chemical potential and temperature T_\nu initially equal to the photon temperature T_\gamma.^[6] Subsequently, as the universe cools to approximately 0.5 MeV, electron-positron pairs annihilate, releasing their entropy primarily into the photon bath via electromagnetic interactions, thereby heating the photons relative to the neutrinos.^[1] This process does not affect the neutrinos, which continue to free-stream without further interactions with the plasma. In the instantaneous approximation, the resulting temperature ratio is T_\nu / T_\gamma = (4/11)^{1/3} \approx 0.714, reflecting the transfer of entropy from two degrees of freedom (e^+ and e^-) to the two photon helicity states.^[6] Non-instantaneous effects during decoupling introduce small spectral distortions, slightly modifying this ratio.^[1] As relativistic particles, the decoupled neutrinos propagate freely, redshifting their momenta with the cosmic expansion such that p \propto 1/a, where a is the scale factor. Their energy density is given by

\rho_\nu = \frac{7}{8} \cdot 6 \cdot \frac{\pi^2}{30} T_\nu^4 = \frac{21}{4} \frac{\pi^2}{30} T_\nu^4,

corresponding to six relativistic degrees of freedom (three flavors, each with left-handed neutrino and right-handed antineutrino).^[6] This contribution is often parameterized relative to the photon energy density via the effective number of neutrino species N_{\rm eff}, defined such that \rho_\nu = N_{\rm eff} \cdot \frac{7}{8} \frac{\pi^2}{30} T_\gamma^4 \left( \frac{T_\nu}{T_\gamma} \right)^4. In the Standard Model, accounting for non-instantaneous decoupling, flavor oscillations, and finite-temperature quantum electrodynamics corrections, N_{\rm eff} \approx 3.044.^[6] Neutrinos maintain their relativistic nature until late cosmic times, with present-day temperatures around 1.95 K (or T_{\nu 0} \approx 1.7 \times 10^{-4} eV), unlike baryonic matter which becomes non-relativistic much earlier due to matter domination.^[1] Their phase-space distributions freeze out in comoving coordinates, preserving the distorted spectra from decoupling and annihilation as relics of the early universe.^[6]

Cosmological implications

Role in Big Bang Nucleosynthesis

Big Bang nucleosynthesis (BBN) occurs at temperatures around T \sim 0.1 MeV, shortly after neutrino decoupling at higher temperatures (T \sim 1 MeV), when neutrinos have ceased interacting with the plasma and contribute to the total energy density as free-streaming relativistic particles.^[7] This neutrino energy density, \rho_\nu, parameterized by the effective number of neutrino species N_{\rm eff} \approx 3.046 in the Standard Model, increases the radiation energy density and thus accelerates the Hubble expansion rate H during BBN.^[8] The faster expansion shortens the time available for weak interactions to maintain equilibrium before the neutron-to-proton freeze-out, indirectly influencing the primordial abundances of light elements. The neutron-to-proton ratio evolves through weak interactions in the early universe, freezing out at T_f \approx 0.8 MeV when the weak interaction rate falls below H. At this point, the ratio is n/p \approx \exp(-\Delta m / T_f) \approx 1/6, where \Delta m = m_n - m_p = 1.293 MeV is the neutron-proton mass difference; pre-decoupling weak rates establish the equilibrium approach, while the post-decoupling neutrino contribution to H hastens the freeze-out.^[8] Between freeze-out and BBN, free neutron decay further reduces the ratio to approximately $1/7, setting the neutron reservoir for subsequent nuclear reactions. This evolution is sensitive to the expansion history shaped by decoupled neutrinos, with deviations in N_{\rm eff} altering T_f and thus the frozen n/p.^[8] The standard N_{\rm eff} = 3.046 predicts light element abundances that align well with observations: the helium-4 mass fraction Y_p \approx 0.25, deuterium-to-hydrogen ratio {\rm D/H} \approx 2.5 \times 10^{-5}, and smaller yields of ^3{\rm He} and ^7{\rm Li}. Higher N_{\rm eff} enhances H, leading to an earlier freeze-out at higher T_f, a larger n/p, increased Y_p, and decreased {\rm D/H} due to more rapid processing of deuterium into helium.^[8] These predictions match primordial abundances inferred from astrophysical measurements, confirming the Standard Model neutrino content. BBN provides stringent constraints on extra relativistic species, such as sterile neutrinos, through sensitivities to \Delta N_{\rm eff} > 0. An additional species would boost H, shifting abundances beyond observed values—for instance, \Delta N_{\rm eff} \gtrsim 1 overproduces Y_p and underproduces {\rm D/H}, yielding upper limits like N_{\rm eff} < 3.5 at 95% confidence from combined BBN data (as of 2024).^[8] This excludes many sterile neutrino models unless their production is suppressed, highlighting BBN's role in testing beyond-Standard-Model physics.

Influence on expansion history

After neutrino decoupling, the relic neutrinos contribute significantly to the total radiation energy density in the early universe, parameterized by the effective number of relativistic neutrino species N_{\rm eff}. In the standard model with three neutrino families, N_{\rm eff} = 3.046, leading to neutrinos accounting for approximately 40% of the radiation density during the radiation-dominated era. This enhanced radiation density increases the Hubble expansion rate H(t) compared to a photon-only scenario, as the Friedmann equation dictates H^2 \propto \rho_{\rm tot}, where \rho_{\rm rad} \propto g_* T^4 and g_* includes the neutrino contribution post-decoupling.^[9]^[10] The increased H(t) delays the epoch of matter-radiation equality, shifting it to a lower redshift z_{\rm eq} \approx 3400 rather than \sim 3600 without neutrinos. At equality, the matter and radiation densities are comparable (\rho_m \approx \rho_{\rm rad}), marking the transition where gravitational instability begins to drive structure growth dominantly. Without this delay, perturbations would enter the horizon earlier, altering the subsequent evolution of cosmic expansion and the timing of later epochs like recombination.^[9] As free-streaming particles, decoupled neutrinos exhibit anisotropic stress that suppresses the growth of density perturbations on small scales (k \gtrsim 0.01 \, h \, \rm Mpc^{-1}) during the radiation era. This free-streaming prevents neutrinos from clustering effectively until they become non-relativistic around z \sim 100, for typical neutrino masses \sum m_\nu \sim 0.06 \, \rm eV. The suppression manifests as a scale-dependent damping in the matter power spectrum, reducing power by a factor of approximately $1 - 8 f_\nu on sub-horizon scales, where f_\nu is the neutrino fraction of total matter density at late times; this effect, akin to Silk damping in the baryon-photon fluid, inhibits early small-scale structure formation until neutrino clustering resumes post-nonrelativistic transition.^[9] Cosmological observations tightly constrain N_{\rm eff} through its imprints on expansion and perturbations. As of 2024, DESI BAO combined with CMB data yield N_{\rm eff} = 3.10 \pm 0.17 (68% CL), in excellent agreement with the standard model prediction and disfavoring significant deviations at high confidence.^[11] Values of N_{\rm eff} > 3.046 would further accelerate early expansion, shifting acoustic peaks in the cosmic microwave background and enhancing small-scale damping, while N_{\rm eff} < 3 could indicate incomplete thermalization or new cooling mechanisms; such deviations probe extensions beyond the standard model, including sterile neutrinos or large extra dimensions that alter the relativistic degrees of freedom.^[9] At late times, the influence of neutrinos on cosmic expansion becomes negligible during the dark energy-dominated acceleration era, as their energy density redshifts as non-relativistic matter (\rho_\nu \propto a^{-3}) and constitutes only \Omega_\nu h^2 \approx \sum m_\nu / (93.14 h^2) \, \rm eV \lesssim 0.0013 for \sum m_\nu < 0.12 \, \rm eV (95% CL upper limit as of 2024).^[9]^[12] However, precise measurements of N_{\rm eff} and neutrino masses remain crucial for forecasting uncertainties in future surveys, such as those probing the growth rate of structure and the Hubble tension, where even small variations in early radiation can propagate to percent-level effects on late-time parameters.^[9]^[11]

Observational evidence

Constraints from primordial abundances

The primordial abundances of light elements, particularly helium-4 and deuterium, provide key observational inputs for testing neutrino decoupling in the context of Big Bang Nucleosynthesis (BBN). The primordial mass fraction of ^4He, denoted Y_p, is measured to be 0.245 ± 0.003 (as of PDG 2024) from observations of metal-poor H II regions in nearby galaxies, where helium lines are analyzed to extrapolate to zero metallicity. A June 2025 measurement from the EMPRESS survey reports a lower value of Y_p = 0.2387^{+0.0036}_{-0.0031}, suggesting better alignment with BBN predictions for the standard model.^[13] Similarly, the deuterium-to-hydrogen ratio D/H is determined to be (2.547 ± 0.029) × 10^{-5} (as of PDG 2024) primarily from absorption features in high-redshift quasar spectra, reflecting the pristine gas in damped Lyman-alpha systems minimally processed by stellar nucleosynthesis. These measurements serve as direct probes of the early universe conditions around neutrino decoupling, as the expansion rate during BBN—governed in part by the neutrino energy density—influences the freeze-out of nuclear reactions. The success of BBN in the 1960s and 1970s, as originally formulated, validated the foundational assumptions of neutrino decoupling and thermal equilibrium in the standard model. Seminal calculations by Peebles (1966) and Wagoner, Fowler, and Hoyle (1967) demonstrated that the observed light element abundances could be reproduced assuming three neutrino species decoupling before BBN, setting the stage for modern constraints. In contemporary analyses, the standard neutrino decoupling scenario yields an effective number of neutrino species N_eff ≈ 3.046, consistent with the three active flavors after accounting for non-instantaneous decoupling effects, while the predicted weak interaction strength parameterized by the Fermi constant G_F agrees with its laboratory-measured value to within 1σ. However, mild tensions arise, particularly with the observed Y_p and deuterium abundances compared to BBN predictions for the Planck-inferred baryon density, potentially indicating astrophysical systematics such as uncertainties in ionization corrections for H II regions or stellar helium production. The 2025 EMPRESS result may help alleviate the Y_p tension. These primordial abundances impose stringent bounds on physics beyond the standard model, particularly extensions involving additional relativistic degrees of freedom. Analyses of D/H and the primordial ^7Li/H ratio yield ΔN_eff < 0.28 at 95% confidence level (from PDG 2024 BBN fits), where ΔN_eff quantifies deviations from the standard N_eff = 3.046, effectively ruling out many sterile neutrino models with masses in the eV to keV range that would enhance the early expansion rate and alter abundances. Updated 2024/2025 BBN calculations incorporating the latest Planck and DESI baryon density further tighten these limits, confirming consistency with standard decoupling while excluding scenarios with extra light species that overproduce helium or underproduce deuterium.^[8]^[14]

Signatures in Cosmic Microwave Background

Neutrino decoupling leaves distinct imprints on the cosmic microwave background (CMB) anisotropies primarily through alterations in the early universe's expansion rate and the dynamics of the baryon-photon fluid after neutrinos begin free-streaming.^[15] Following decoupling around 1 MeV, the relativistic neutrinos contribute to the radiation density without participating in pressure-supporting oscillations with photons and baryons, effectively increasing the Hubble expansion and modifying the sound horizon.^[9] This leads to observable effects in the CMB power spectrum, including enhanced damping at small scales and shifts in the positions of acoustic peaks, providing indirect probes of the standard effective number of neutrino species, N_\mathrm{eff} = 3.046.^[16] One key signature is Silk damping, where free-streaming neutrinos influence the diffusion of photons in the baryon-photon plasma, suppressing power on small angular scales. The sound speed of this fluid is given by

c_s^2 = \frac{1}{3(1 + R)},

where R \propto \rho_b / \rho_\gamma measures the baryon-to-photon density ratio; the presence of free-streaming neutrinos reduces the effective c_s by accelerating expansion without adding to the fluid's pressure, enhancing photon random walks and damping fluctuations for wavenumbers k > 0.02 \, \mathrm{Mpc}^{-1}.^[15] Higher N_\mathrm{eff} amplifies this effect, increasing the damping scale and lowering the high-\ell tail of the temperature power spectrum, as the additional relativistic degrees of freedom hasten diffusion before recombination.^[9] Neutrino decoupling also induces a phase shift in the acoustic peaks of the CMB temperature (TT) spectrum, altering the odd-even parity due to the altered timing of photon-baryon oscillations from faster early expansion. This shift is approximated as \Delta \theta \approx -0.4 (N_\mathrm{eff} - 3)/3.046, moving peaks toward higher multipoles \ell for N_\mathrm{eff} > 3.046; observations of the Planck TT spectrum confirm consistency with the standard value, with deviations constrained to less than 0.2 in phase.^[16] The effect arises because free-streaming neutrinos drag the gravitational potentials less effectively than a tightly coupled fluid, shifting the oscillation phase without strongly affecting peak amplitudes.^[15] At low multipoles (\ell \lesssim 10), neutrino decoupling impacts the early integrated Sachs-Wolfe (ISW) effect, where evolving potentials during the radiation-matter transition imprint on CMB polarization. Relativistic neutrinos enhance metric perturbations at early times, boosting the E-mode polarization power spectrum through this ISW contribution and providing constraints on N_\mathrm{eff} from low-\ell TE and EE correlations.^[9] Standard decoupling yields a modest early ISW signal consistent with Planck polarization data, suppressing power slightly compared to models with fewer neutrinos.^[16] Recent analyses combining Planck and Atacama Cosmology Telescope (ACT) DR6 data reaffirm these signatures, yielding N_\mathrm{eff} = 2.99 \pm 0.17 (Planck 2018) and N_\mathrm{eff} = 2.86 \pm 0.13 for ACT DR6 (2025), with combined constraints tightening ΔN_eff < 0.17 at 95% CL and no significant deviations from standard decoupling.^[16]^[17] These measurements update indirect evidence from phase shifts and damping, confirming neutrino free-streaming as the dominant mechanism without requiring extensions beyond the standard model.^[17]