The first search for extremely-high energy cosmogenic neutrinos with the IceCube Neutrino Observatory

We report on the results of the search for extremely-high energy (EHE) neutrinos with energies above $10^7$ GeV obtained with the partially ($\sim$30%) constructed IceCube in 2007. From the absence of signal events in the sample of 242.1 days of effective livetime, we derive a 90% C.L. model independent differential upper limit based on the number of signal events per energy decade at $E^2 \phi_{\nu_e+\nu_\mu+\nu_\tau}\simeq 1.4 \times 10^{-6}$ GeV cm$^{-2}$ sec$^{-1}$ sr$^{-1}$ for neutrinos in the energy range from $3\times10^7$ to $3\times10^9$ GeV.

💡 Research Summary

The paper reports the first search for extremely‑high‑energy (EHE) cosmogenic neutrinos using the IceCube Neutrino Observatory during its early construction phase in 2007, when only about 30 % of the detector was operational. Cosmogenic neutrinos are expected to be produced when ultra‑high‑energy cosmic rays interact with the cosmic microwave background, generating neutrinos with energies above 10⁷ GeV (10 PeV). Detecting such particles would provide a unique probe of the sources of the highest‑energy cosmic rays and of particle interactions at energies far beyond the reach of terrestrial accelerators.

IceCube consists of an array of Digital Optical Modules (DOMs) embedded in the Antarctic ice, each recording Cherenkov photons emitted by charged particles produced in neutrino‑nucleon interactions. In 2007 the detector comprised 86 strings with 5,160 DOMs, but only a subset (≈30 %) was active. Data corresponding to 242.1 days of effective livetime were collected, yielding roughly 10⁸ triggered events, the vast majority of which are atmospheric muons and neutrinos.

To isolate a potential EHE neutrino signal, the analysis employed a two‑stage filtering strategy. The first stage selected events with large total charge (≥1000 photo‑electrons) and characteristic waveforms indicative of high‑energy cascades or muon tracks. The second stage reconstructed the event topology and arrival direction, focusing on near‑horizontal trajectories (zenith angles ≈80°–100°) where the atmospheric muon background is most suppressed. Simulations of atmospheric muons (CORSIKA) and atmospheric neutrinos (ANIS) were used to estimate the residual background, while signal expectations were generated from several cosmogenic neutrino models (e.g., Engel‑Seckel‑Stanev, Kotera‑Allard‑Murase).

The effective area of the partially built detector rises steeply with energy, reaching ~0.5 km² at 10⁸ GeV for horizontal events. Systematic uncertainties arise from ice optical properties (absorption and scattering lengths), DOM sensitivity, the modeling of the atmospheric muon flux, and neutrino‑nucleon cross‑sections. Each source contributes roughly 10–20 % uncertainty, leading to an overall systematic error of about 30 % on the derived flux limit.

No candidate EHE neutrino events survived the final selection. Consequently, the authors set a 90 % confidence‑level, model‑independent differential upper limit on the all‑flavor neutrino flux:

\