Search for nucleon decay via $p ightarrowνπ^{+}$ and $n ightarrowνπ^{0}$ in 0.484 Mton-year of Super-Kamiokande data

We present the results of searches for nucleon decays via $p\rightarrowνπ^{+}$ and $n\rightarrowνπ^{0}$ using a 0.484 Mt$\cdot$yr exposure of Super-Kamiokande I-V data covering the entire pure water phase of the experiment. Various improvements on the previous 2014 nucleon decay search, which used an exposure of 0.173 Mt$\cdot$yr, are incorporated. The physics models related to pion production and nuclear interaction are refined with external data, and a more comprehensive set of systematic uncertainties, now including those associated with the atmospheric neutrino flux and pion production channels is considered. Also, the fiducial volume has been expanded by 21%. No significant indication of a nucleon decay signal is found beyond the expected background. Lower bounds on the nucleon partial lifetimes are determined to be $3.5\times10^{32}$~yr for $p\rightarrowνπ^{+}$ and $1.4\times10^{33}$~yr for $n\rightarrowνπ^{0}$ at 90% confidence level.

💡 Research Summary

The paper reports a comprehensive search for the nucleon decay modes p → ν π⁺ and n → ν π⁰ using the full pure‑water data set of the Super‑Kamiokande (SK) detector spanning phases I through V. The total exposure amounts to 0.484 megaton·years (Mt·yr), corresponding to 17.8 years of live time, which is a 2.8‑fold increase over the previous SK nucleon‑decay analysis published in 2014 (0.173 Mt·yr). The authors implement several key improvements: (1) inclusion of the SK‑IV and SK‑V data, raising the livetime by 132 %; (2) expansion of the fiducial volume by 21 % (from the traditional 22.5 kton to ≈27.2 kton), thereby increasing the number of target nucleons; (3) refinement of the physics models governing pion production and nuclear interactions using recent external measurements, especially updated π‑N scattering data and spectral‑function based descriptions of bound nucleons; and (4) incorporation of additional systematic uncertainties related to the atmospheric neutrino flux and neutrino‑nucleon interaction modeling, which were not fully accounted for in the earlier analysis.

Signal Monte‑Carlo (MC) events are generated separately for protons bound in oxygen and free protons in hydrogen (for p → ν π⁺) and for neutrons bound in oxygen (for n → ν π⁰). Bound nucleons are assigned to s‑ or p‑shell states according to the nuclear shell model (25 % s‑state, 75 % p‑state) and are given momenta drawn from a spectral function measured in electron‑carbon scattering experiments. Nuclear binding energies are modeled with Gaussian distributions (mean 39 MeV, σ = 10.2 MeV for s‑state; mean 15.5 MeV, σ = 3.8 MeV for p‑state). Ten percent of bound‑nucleon decays are treated as “correlated decays,” where the surrounding nucleons broaden the kinematic distributions of the emitted pion.

Backgrounds arise from atmospheric neutrino interactions simulated with the NEUT generator, using the Honda‑Kajita‑Kasahara‑Midorikawa flux model and three‑flavor oscillation parameters. The dominant background channels are charged‑current quasielastic (CCQE) scattering and neutral‑current (NC) single‑pion production. The authors also include two‑nucleon knockout (2p‑2h) processes and employ the Rein‑Sehgal model for resonant pion production, tuned to recent SK NC π⁰ control samples. Differences between the signal and background nuclear models (e.g., spectral‑function vs. Fermi‑gas) are treated as systematic uncertainties.

Event selection focuses on single‑ring Cherenkov topologies. For the π⁰ channel, two‑photon rings are reconstructed and required to form an invariant mass consistent with the π⁰ mass; for the π⁺ channel, a single ring with an associated Michel electron (from μ → e decay) is required. Additional cuts on reconstructed momentum (100 MeV–1.5 GeV) and fiducial volume are applied. The resulting selection efficiencies are ≈45 % for p → ν π⁺ and ≈38 % for n → ν π⁰. Expected background yields after all cuts are 0.96 events for the charged‑pion mode and 1.12 events for the neutral‑pion mode.

A Bayesian statistical analysis with Poisson likelihoods, incorporating all systematic uncertainties (detector response, flux, cross‑sections, nuclear modeling), is performed to set limits on the partial lifetimes. No excess over the predicted background is observed. Consequently, the authors derive 90 % confidence level lower limits of τ/B > 3.5 × 10³² yr for p → ν π⁺ and τ/B > 1.4 × 10³³ yr for n → ν π⁰. These limits improve upon the previous SK results (3.9 × 10³² yr and 1.1 × 10³³ yr, respectively) by roughly 10 % and 27 %.

The paper discusses the implications for Grand Unified Theories. In minimal supersymmetric SO(10) models with a 126‑dimensional Higgs field that breaks B − L symmetry, the p → ν π⁺ and n → ν π⁰ modes can dominate in certain parameter regions. The new limits therefore exclude a substantial portion of the viable parameter space for such models. The authors also note that the methodological advances—larger exposure, expanded fiducial volume, refined nuclear modeling, and comprehensive systematics—provide a robust framework for future nucleon‑decay searches in upcoming detectors such as Hyper‑Kamiokande, DUNE, and the gadolinium‑doped SK‑II phase.

In summary, the analysis delivers the most stringent constraints to date on the p → ν π⁺ and n → ν π⁰ decay channels, demonstrates the power of incremental improvements in detector operation and modeling, and sets a benchmark for the next generation of proton‑decay experiments seeking to probe physics at the grand‑unification scale.