Updated Constraints on Large Extra Dimensions from Reactor Antineutrino Experiments

TL;DR

This paper updates constraints on large extra dimensions (LED) using comprehensive reactor antineutrino data, extending the analysis to both a single LED ($d=1$) and equal-radius $4+d$ configurations with up to four extra dimensions. The authors implement a full KK-mode mixing framework in a $4+d$ neutrino sector, deriving the oscillation probabilities and incorporating them into reactor flux simulations across Daya Bay, RENO, KamLAND, NEOS, and STEREO. They report improved bounds on the LED radius $a$, with NO and IO bounds tightening as the number of extra dimensions increases, and demonstrate that Daya Bay largely drives the constraints. The results illustrate the continued power of precision reactor neutrino experiments to probe physics beyond the Standard Model, including potential explanations for neutrino mass generation via extra dimensions, and set the stage for future, more sensitive measurements.

Abstract

We investigate constraints on large extra dimensions (LED) using the latest results from reactor antineutrino experiments. Specifically, we analyze the full data sets from Daya Bay, RENO, KamLAND, NEOS, and STEREO to derive updated bounds. For the case of one extra dimension, we find constrains on its radius $a$ of $a \lesssim 0.58~{\rm μm}$ ($a \lesssim 0.12~{\rm μm}$) at the $99\%$ confidence level for normal (inverted) ordering, an improvement of approximately $\sim 20\%$ ($\sim 25\%$) with respect to previous bounds, assuming a massless lightest active neutrino. Furthermore, we present new limits on $4+d$ LED scenarios, with $d = 2, 3, 4$ denoting the number of extra dimensions, based on the same reactor data and assuming equal radii for all extra dimensions. We find that the constraints become increasingly stringent with a larger number of extra dimensions. In particular, $d = 4$ with a massless lightest active neutrino, we obtain limits of $a \lesssim 0.28~{\rm μm}$ for normal and $a \lesssim 0.05~{\rm μm}$ for inverted orderings at the $99\%$ confidence level.

Figures (5)

• Figure 1: Electron antineutrino survival probability as a function of energy in the presence of LED, shown for $d = 1$ (purple), $d = 2$ (orange), $d = 3$ (green), and $d=4$ (turquoise). Each row corresponds to a different baseline: $L = 24$ m (top), $L = 1$ km (middle), and $L = 180$ km (bottom). Left and right panels show results for normal and inverted mass orderings, respectively. The parameters are fixed to $a = 1~\mu\mathrm{m}$ and $m_0 = 0.01$ eV.
• Figure 2: Observed event spectra at Daya Bay, along with the best-fit predictions for the standard 3 $\nu$ framework and the LED model. The standard oscillation parameters used are $\Delta m^2_{31} = 2.5 \times 10^{-3}~\text{eV}^2$ and $\sin^2 2\theta_{13} = 0.085$, while the LED parameters are $a = 3~\mu\text{m}$ and $m_0 = 0.1~\text{eV}$. The event distribution is shown as a function of the positron kinetic energy.
• Figure 3: Large extra dimension constraints $a$ and ${\rm m}_0$ for $d=1$, i.e., 5D for the NO (left) and IO (right). We present the excluded regions from Daya Bay (orange), RENO (blue), KamLAND (dark blue), STEREO (green), NEOS (fuchsia) and their combination (black) at 99% (full) C.L.
• Figure 4: Large extra dimension constraints $a$ and ${\rm m}_0$ for the NO (left) and IO (right). We present the excluded regions for $d=1$ (purple), $d=2$ (orange), $d=3$ (green) and $d=4$ (turquoise) at 99% C.L.
• Figure 5: Analysis of Daya Bay (left) and RENO (right) final data release considering the $3\nu$ scenario. The results of this analysis are given by the colored regions, where each region corresponds to the $1\sigma,2\sigma$ and $3\sigma$. See the text for details about the analysis. The results are compared with Daya Bay DayaBay:2022orm and RENO RENO:2024msr results shown by the dashed lines.