Constraints on secret neutrino interactions after Planck

TL;DR

This work probes hidden neutrino interactions mediated by a light (pseudo)scalar, such as Majoron models, by modifying the Boltzmann evolution of neutrinos and fitting to CMB data from Planck, ACT, SPT, and BK. Using a relaxation-time collision term in the Boltzmann hierarchy, the authors constrain the coupling strength via the parameter γ_{ u u}^4, translating into limits on the recoupling redshift z_{ u rec} and, in Majoron scenarios, the coupling g. Across minimal and extended ΛCDM frameworks, they find γ_{ u u}^4 ≲ (0.9–1.0)×10^{-27} (95% C.L.), with best-fit hints around (0.15–0.35)×10^{-27}, suggesting potential but not definitive evidence for late-time recoupling. The results imply z_{ u rec} up to about 8500 and map to Majoron-neutrino couplings g ≲ 8×10^{-7}, offering meaningful constraints on beyond-Standard-Model neutrino interactions and guiding future cosmological probes.

Abstract

(Abridged) Neutrino interactions beyond the standard model may affect the cosmological evolution and can be constrained through observations. We consider the possibility that neutrinos possess secret scalar or pseudoscalar interactions mediated by the Nambu-Goldstone boson of a still unknown spontaneously broken global $U(1)$ symmetry, as in, e.g. , Majoron models. In such scenarios, neutrinos still decouple at $T\simeq 1$ MeV, but become tightly coupled again ('recouple') at later stages of the cosmological evolution. We use available observations of CMB anisotropies, including Planck 2013 and the joint BICEP2/Planck 2015 data, to derive constraints on the quantity $γ_{νν}^4$, parameterizing the neutrino collision rate due to (pseudo)scalar interactions. We consider both a minimal extension of the standard $Λ$CDM model, and scenarios with extra relativistic species or non-vanishing tensors. We find a typical constraint $γ_{νν}^4 < 0.9\times 10^{-27}$ (95% C.L.), implying an upper limit on the redshift $z_{rec}$ of neutrino recoupling $< 8500$. In the framework of Majoron models, the upper limit on $γ_{νν}$ roughly translates on a constraint $g < 8.2\times 10^{-7}$ on the Majoron-neutrino coupling constant $g$. In general, the data show a weak ($\sim 1σ$) but intriguing preference for non-zero values of $γ_{νν}^4$, with best fits in the range $γ_{νν}^4 = (0.15 - 0.35)\times 10^{-27}$, depending on the particular dataset. This is more evident when either observations from ACT and SPT are included, or the possibility of non-vanishing tensor modes is considered. In particular, for the minimal model $Λ$CDM +$γ_{νν}$ and including the Planck 2013, ACT and SPT data, we report $γ_{νν}^4=( 0.45^{+0.15}_{-0.38} )\times10^{-27}$ ($200 < z_{rec} < 5700$) at 68% confidence level.

Figures (7)

• Figure 1: Feynman diagrams for the binary processes allowed by the Lagrangian (\ref{['eq:lagr']}). Time goes from left to right. From left to right: $\nu$-$\nu$ scattering (s and t channels), $\nu$-$\phi$ scattering, $\nu\bar{\nu}$ annihilation to $\phi$'s.
• Figure 2: Evolution of neutrino density (left column) and shear (right column) perturbations for three different modes $k = 0.005,\, 0.05,\, 0.5\, \mathrm{Mpc}^{-1}$. We compare the evolution for standard, weakly-interacting neutrinos (blue solid curves) with the case of non-standard neutrino interactions with $\gamma_{\nu\nu} =1.2 \times 10^{-7}$ and $\gamma_{\nu\nu} = 2 \times 10^{-7}$ (green dashed and red dotted curves respectively). Non-standard interactions suppress shear perturbations while boosting density perturbations after the time of neutrino recoupling (see text for discussion).
• Figure 3: CMB temperature (left panel) and temperature-$E$ polarization (right panel) power spectra for neutrinos with non-standard interactions. As in Fig. \ref{['fig:perturb-evolution']} we show results for $\gamma_{\nu\nu} = \{0,\, 1.2 \times 10^{-7},\,2 \times 10^{-7}\}$.
• Figure 4: One-dimensional posterior distribution for $\gamma_{\nu\nu}^4$ obtained from the Planck+WP (blue solid) and Planck+WP+highL (red dashed) datasets. The shaded regions denote the 68% credible interval. Left panel: $\Lambda$CDM + $\gamma_{\nu\nu}$. Right panel: $\Lambda$CDM+$\gamma_{\nu\nu}$+$N_\mathrm{eff}$.
• Figure 5: 68% and 95% confidence regions for selected parameter pairs involving $\gamma_{\nu\nu}^4$ in the $\Lambda$CDM+$\gamma_{\nu\nu}$ (empty contours) and $\Lambda$CDM+$\gamma_{\nu\nu}+N_\mathrm{eff}$ (filled contours), for Planck+WP (blue) and Planck+WP+HL (red).