New Constraint on the Local Relic Neutrino Background Overdensity with the First KATRIN Data Runs

We report on the direct search for cosmic relic neutrinos using data acquired during the first two science campaigns of the KATRIN experiment in 2019. Beta-decay electrons from a high-purity molecular tritium gas source are analyzed by a high-resolution MAC-E filter around the end point at 18.57 keV. The analysis is sensitive to a local relic neutrino overdensity ratio of $\eta <9.7\times {10}^{10}/\alpha$ ($1.1\times {10}^{11}/\alpha$) at a 90% (95%) confidence level with $\alpha =1$ (0.5) for Majorana (Dirac) neutrinos. A fit of the integrated electron spectrum over a narrow interval around the end point accounting for relic neutrino captures in the tritium source reveals no significant overdensity. This work improves the results obtained by the previous neutrino mass experiments at Los Alamos and Troitsk. We furthermore update the projected final sensitivity of the KATRIN experiment to $\eta <1\times {10}^{10}/\alpha$ at 90% confidence level, by relying on updated operational conditions.

Figure 1

The main components of KATRIN: (a) the rear section, (b) the windowless gaseous tritium source WGTS, (c) the pumping section, and a tandem setup of two MAC-E- filters: (d) the smaller prespectrometer and (e) the larger main spectrometer with surrounding aircoil system. This setup allows only the highest energy $\beta$-decay electrons to reach the focal plane detector (f) where they are counted.

Figure 2

Top panel: Uniform electron spectra $R(⟨qU⟩)$ of both datasets, with $1\sigma$ error enlarged by a factor of 50. The blue and green lines represent the $\beta$-decay best-fit models $R_{\beta ,\mathrm{calc}}(⟨qU⟩)$. Below $E_0$, the KNM2 spectrum shows higher rates due to the increase in tritium activity from KNM1 to KNM2. Above $E_0$, we note a flat background $R_{\mathrm{bg}}$ reduced by 25% between KNM1 and KNM2. Lower panel: integral distributions of the measurement time of the two datasets.

Figure 3

Relic neutrino differential signal with arbitrary exposure, the detailed model (blue), the Gaussian simplified model (red), and $m_{\nu }=0.7\mathrm{eV}$.

Figure 4

Simulated impact of the molecular effects (FSD) on the $\beta$-decay and neutrino capture differential spectra of molecular tritium and $m_{\nu }=0.7\mathrm{eV}$ (arbitrary exposure). All spectra also include Doppler broadening.

Figure 5

Spectral ratio of the best fits for relic neutrinos with respect to the null hypothesis, $\eta =0$, (top) for the first and (bottom) for the second measurement campaigns. The values of the best fits of $\eta \alpha$ are both consistent with background fluctuations.

Figure 6

Exclusion contours on the relic neutrino overdensity $\eta \alpha$ at 99% C.L., for both measurement campaigns and their combination, and the projected KATRIN sensitivity for each neutrino mass.

Figure 7

Overview of the new 95% C.L. KATRIN limits on the local relic neutrino overdensity ratio, compared to constraints from previous experiments at Los Alamos [8] and Troitsk [7] for which the assumed neutrino model has not been published.