Dark Neutrino interactions phase out the Hubble tension

TL;DR

The paper proposes Dark Neutrino Interactions (DNI) as a mechanism to alleviate the Hubble tension by stopping neutrinos from free streaming, thereby inducing a negative, scale-dependent phase shift that can offset the standard neutrino-induced phase and raise $H_0$ without changing $N_{\rm eff}$. DNI relies on a small fraction $f$ of interacting dark matter $\chi$ coupled to neutrinos via a messenger $\psi$ through a temperature-independent cross-section, producing coupled perturbations that modify the acoustic peaks and the matter power spectrum. Using Planck 2015 data and WiggleZ, with SH0ES, the analysis finds that nonzero DNI interactions are preferred and can reduce the tension from $\sim3.8\sigma$ in $\Lambda$CDM to roughly $2.1$–$2.9\sigma$, while predicting observable effects in CMB $B$-modes and large-scale structure for future confirmation.

Abstract

New interactions of neutrinos can stop them from free streaming even after the weak interaction freeze-out. This results in a phase shift in the cosmic microwave background (CMB) acoustic peaks which can alleviate the Hubble tension. In addition, the perturbations in neutrinos do not decay away on horizon entry and contribute to metric perturbation enhancing the matter power spectrum. We demonstrate that this acoustic phase shift can be achieved using new interactions of standard left-handed neutrinos with dark matter without changing the number of effective relativistic degrees of freedom. Using Planck CMB and the WiggleZ galaxy survey $ (k\le 0.12 h \ {\rm Mpc}^{-1} ) $ data, we demonstrate that in this model the Hubble tension reduces to approximately $ 2.1 σ$. Our model predicts potentially observable modifications of the CMB B-modes and the matter power spectrum that can be observed in future data sets.

Figures (7)

• Figure 1: CMB temperature (TT) power spectrum around first six acoustic peaks. The leftmost solid red line is the best-fit Planck Aghanim:2015xee temperature power spectrum with a best-fit value of $H_0=67.9~{\rm kms^{-1}Mpc^{-1}}$. Introducing DNI, keeping all other cosmological parameters fixed, moves all peaks to the right/higher $\ell$ with larger shift for higher $\ell$ peaks (rightmost solid blue curves). However, DNI with higher $H_0$ brings the peaks back to the original positions (dashed blue). The amplitudes of DNI power spectra for each peak is adjusted so that the peak height is the same as the $\Lambda$CDM. Also shown as points with error bars is the binned Planck power spectrum.
• Figure 2: Comparison of optical depth of neutrinos in DNI $(fu = 0.034)$ with models of neutrino self-interaction Oldengott:2017fhy and 2019arXiv190407810F having different temperature dependences. The top axis shows the modes $\ell_H$ which enter horizon at redshift $z$.
• Figure 3: Shift of the position of peaks of CMB TT $(\Delta\ell_{TT})$and EE $(\Delta\ell_{EE})$ spectrum in $\Lambda$CDM and DNI cosmologies $(f = 10^{-3}, u=34)$ with respect to bestfit $\Lambda$CDM model with $H_0 = 67.9 ~\rm km/s/Mpc$.
• Figure 4: The left panel shows $1\sigma,2\sigma,~\text{and}~3\sigma$ constraints in DNI and $H_0$ for different data set combinations. The light and dark grey band shows $1\sigma$ and $2\sigma$ band for SH0ES Riess:2019cxk measurement respectively. The central panel shows the MCMC samples in the $f-fu$ plane. The right most panel shows calculation of Hubble tension (values given in the legend) taking into account non-Gaussianity of PDFs. The $2\sigma$ upper-limit from P15 is $fu <0.034$.
• Figure 5: Left: 1-D posterior for $fu$ for fixed $f$ DNI cosmology (see Table. \ref{['tbl:chisq']}). The dashed vertical lines marks the $3\,\sigma$ upper and lower limit for the corresponding dataset. Right: The $x\,\sigma$ lower limit on $fu$ plotted against $x$. It can be seen from both the plot that $fu = 0$ is excluded at more than $3\,\sigma$ when we include SH0ES data.