Constraining the cosmic radiation density due to lepton number with Big Bang Nucleosynthesis

This is the paper that turns primordial neutrino asymmetries into a decision rule for Planck's N_eff, and it does that job well enough that I'd cite it. The one caveat to carry in your head: the headline 3.4 is read off the boundary of the allowed region, where the authors themselves report non-thermal distortions worth ΔN_eff ≈ 0.2, and those distortions are not in the weak-rate integrals that set the boundary. More on that below, but it is a precision issue, not a refutation. The qualitative statement — for sin²θ13 ≳ 10⁻³ the bound collapses to the standard 3.046, for smaller θ13 it relaxes to ~3.4 — is robust and is the useful part for Planck interpretation. What's genuinely new relative to Pastor–Pinto–Raffelt (their ref. 21): a wide scan of initial asymmetries, the evolved distributions fed into a modified PArthENoPE, and the θ13 dependence in Figure 9. The pipeline is clean — kinetic equations forward from 10 MeV, distributions into a public BBN code, comparison to external D/H and Yp, nothing tuned to the abundances being predicted. Carrying two helium determinations in parallel is the right call given how contested that measurement is, and Table 1 plus Figures 5–8 are the useful deliverables. The soft spot is the two-parameter Fermi–Dirac refit in Section 3. The Hubble rate uses the actual energy density; the n↔p rates use the fitted FD. At the boundary point (η_νe^fin = 0, η_νx^fin = 0.3) they quote true N_eff = 3.4 against 3.2 from the equilibrium formula — that gap is exactly the shape distortion, and it is absent from the weak-rate integrals. They never quantify the resulting shift in Yp or D/H. So read 3.4 as approximate, maybe ±0.1, not a hard number. Minor: the deuterium error is inflated by √3.6 ≈ 1.9, which weakens the constraint the paper itself calls decisive; and the kinetic code isn't public, only PArthENoPE. I disagree with one reader point: Figure 9 is a scan over θ13 down to 10⁻⁴, not an interpolation between two points, and Figure 2 does show the real-vs-FD comparison directly — the issue is the split use of the two descriptions, not missing validation. Send it to referees. JCAP-appropriate, and I'd ask specifically for an estimate of the Yp shift from the FD refit at the boundary.

The pipeline uses two different descriptions of the same neutrino ensemble at the same epoch. The Hubble rate is modified 'to account for the actual evolution of total neutrino energy density' (Section 3), i.e. the true, non-thermal energy density that yields N_eff = 3.4. But the weak rates that set neutron-proton chemical equilibrium are 'averaged over the corresponding electron (anti)neutrino distribution' after fitting the true spectra to a Fermi-Dirac form with two parameters, T_να and ξ_να, chosen to reproduce the first two moments (asymmetry and energy density). A two-moment fit preserves number and energy but not shape; the n↔p rates depend on the ν_e and ν̄_e spectra across the ~1.3 MeV threshold, so a shape distortion that leaves the moments fixed can still shift Yp. The paper's justification — 'for all initial values of neutrino asymmetries which are compatible with BBN bounds, the distortions ... are typically quite small' — is validated only inside the region the analysis is trying to constrain, and Figure 2 shows the real ν_e spectrum visibly departing from the fitted curve. The tension is sharpest exactly where the headline number lives: the authors note that at the boundary point (η_ν_e^fin = 0, η_ν_x^fin = 0.3) the true N_eff is 3.4 while the equilibrium formula gives 3.2, attributing the difference to non-thermal distortion. That same distortion is absent from the weak-rate integrals that define the boundary. The paper does not quantify the resulting error in Yp or D/H, so the precision of the quoted N_eff = 3.4 is not established.

The reader's weakest_assumption correctly identifies the two-parameter Fermi-Dirac refit as the soft spot, and I am sharpening rather than replacing it: the issue is not merely that the refit is approximate, but that the paper uses the true distribution for the energy density (giving N_eff = 3.4) while using the refit for the weak rates (giving the BBN bound). This is an internal inconsistency at precisely the point where the headline number is extracted, and the authors' own remark that the boundary point shows ΔN_eff ≈ 0.2 of non-thermal origin confirms the distortion is non-negligible there. The paper does acknowledge this non-thermal contribution, so I am not presenting an unacknowledged limitation as new; what remains unresolved is that the acknowledgment is not propagated into an error estimate on Yp or on the quoted N_eff. That said, the concern bears on the precision of the number, not on the existence or direction of the bound: the qualitative result that primordial asymmetries can push N_eff above 3.046 up to roughly 3.4 for small θ13, and the Planck decision rule that a measured non-zero θ13 would rule out asymmetries as the source of an N_eff excess, both survive. I therefore recommend UNCHANGED rather than CONDITIONAL. The other soft points the reader raises are lesser: the D/H error inflation by √(χ²_min/6) ≈ 1.9 makes the bound conservative (a tighter D/H would lower, not raise, the allowed N_eff), so it cannot threaten the claim; and the Figure 9 interpolation between only two computed θ13 values affects the intermediate region, not the θ13 → 0 endpoint where the headline bound sits. The paper's honesty in presenting two Yp determinations and in flagging extra relativistic species as a loosening factor is genuine independent support for the robustness of the qualitative conclusion. The one concrete check above would settle whether the quoted 3.4 is controlled at the stated precision.

The analysis is a genuine forward calculation: the asymmetries are scanned as inputs, the kinetic equations are integrated, and the nucleosynthesis network is an independent public code, so nothing is fitted to the abundances being predicted. The use of two 4He determinations that bracket the current systematic spread, and the demonstration that deuterium breaks the ην–ηνein degeneracy left by 4He alone (Figure 6), are well motivated. The result is falsifiable: the predicted N_eff ceiling as a function of θ13 (Figure 9) can be confronted with the forthcoming Planck measurement, and the paper states clearly what a Planck value above 4 would imply. The finding that ηνxfin − ηνefin can be of order 0.3 for θ13 = 0 is a useful correction to a standard simplifying assumption in the BBN literature.

The paper asks how much primordial neutrino–antineutrino asymmetry can survive flavor oscillations and still agree with the measured primordial deuterium and helium abundances, and what that implies for the radiation density of the early universe. It evolves the three active neutrino distributions with density-matrix kinetic equations that include collisions, pair processes, and oscillations, then feeds the resulting spectra into a Big Bang nucleosynthesis code. The payoff is a set of 95% confidence bounds on the total neutrino lepton asymmetry $eta_ν$ and the initial electron-neutrino asymmetry $eta_{ν_e}^{in}$, with the deuterium abundance breaking a degeneracy that helium alone leaves open. The central number is that, for $sin^2θ_{13} lesssim 10^{-3}$, primordial asymmetries can push the effective number of neutrinos $N_{eff}$ up to about 3.4, whereas for $sin^2θ_{13}=0.04$ the bound collapses close to the standard 3.046; a Planck measurement of $N_{eff}$ can therefore help separate lepton asymmetry from extra relativistic species.

What carries the argument. The argument is carried by a $3×3$ density-matrix kinetic equation for each neutrino momentum, $i,dvarrho_p/dt=[Ω_p,varrho_p]+C[varrho_p,barvarrho_p]$, in which the oscillation term includes mass differences and matter effects and the collision term includes scattering and pair processes. The evolved distributions are then compressed into Fermi-Dirac forms with an effective temperature $T_ν(T_γ)$ and effective degeneracy $ξ_ν(T_γ)$; those two moments enter the weak rates and the Hubble rate in a modified PArthENoPE BBN calculation. This is the mechanism that lets the paper connect initial flavor asymmetries to the final helium and deuterium yields rather than treating $N_{eff}$ and the chemical potentials as fixed parameters.

Core claim. The paper claims that Big Bang nucleosynthesis, combined with the WMAP baryon density and the measured primordial abundances of deuterium and helium-4, restricts the total neutrino lepton asymmetry and the initial electron-neutrino asymmetry strongly enough to bound the radiation contribution of primordial asymmetries. For $θ_{13}=0$ and $Y_p=0.250±0.003$ it finds $-0.66<eta_ν<0.63$ at 95% C.L.; for $sin^2θ_{13}=0.04$ the range shrinks to $-0.13<eta_ν<0.07$. The corresponding final electron-neutrino asymmetry is constrained to $-0.03 lesssim eta_{ν_e}^{fin} lesssim 0.02$ for the lower helium value, while the difference between final $ν_x$ and $ν_e$ asymmetries can reach about $±0.3$ when $θ_{13}=0$ but only about $±0.04$ when $sin^2θ_{13}=0.04$. Translating these ranges into radiation density gives $N_{eff} lesssim 3.4$ for small $θ_{13}$ and a value near the standard 3.046 for the larger mixing.

The premise it rests on. The load-bearing premise is that the evolved neutrino spectra can be represented by Fermi-Dirac functions with only an effective temperature and chemical potential, so that residual non-thermal distortions do not materially change the weak rates or the expansion rate.

For $sin^2θ_{13} lesssim 10^{-3}$, a Planck value of $N_{eff}$ between 3.046 and about 3.4 could be explained by primordial neutrino asymmetry alone, without new particles.

For $sin^2θ_{13} simeq 0.04$, the same asymmetries are forced to be small, so any Planck excess above 3.046 would point to relativistic degrees of freedom beyond the three active neutrinos.

The deuterium abundance, combined with the WMAP baryon density, is what closes the allowed region; using helium-4 alone leaves a degeneracy between $eta_ν$ and $eta_{ν_e}^{in}$.

When $θ_{13}=0$, the final $ν_x$ and $ν_e$ asymmetries need not be equal, so the common assumption of full flavor equilibration before BBN is not required by the data.

A future measurement of $θ_{13}$ together with Planck's $N_{eff}$ can decide whether an excess radiation density is leptonic or due to extra species.

If Planck finds $N_{eff} approx 3.2$–$3.4$ and oscillation experiments find $θ_{13}$ near zero, the simplest reading would be a cosmic neutrino lepton asymmetry rather than a new light particle; if $θ_{13}$ is large, the same $N_{eff}$ would require extra radiation.

The two-moment Fermi-Dirac compression suggests a direct check: rerun the BBN code with the full spectra and compare the contour shifts. A null shift would make the $sim 3.4$ ceiling hard; a visible shift would make it soft.

Because the analysis treats extra radiation and neutrino asymmetry one at a time, a joint likelihood over $N_{eff}$, $eta_ν$, and $θ_{13}$ would be more informative than either bound alone, especially if both effects are present.

The allowed non-equilibrium splitting between $ν_x$ and $ν_e$ for $θ_{13}=0$ leaves a possible target for future neutrino-spectrum or CMB distortion searches, though the paper does not compute such a signature.

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review