Simulation of neutrons produced by high-energy muons underground

TL;DR

This paper uses Monte Carlo simulations with GEANT4 to estimate muon-induced neutron flux underground, comparing results with measurements and identifying over-production in lead, relevant for dark matter experiments.

Contribution

It provides a validated simulation framework for muon-induced neutrons underground and quantifies neutron yield in lead at high energies, aiding background estimation for dark matter searches.

Findings

Simulation overestimates neutron rate by 1.8 times in lead.

Estimated neutron yield in lead is (1.31 +/- 0.06) x 10^(-3) neutrons/muon/(g/cm^2).

No neutron-induced events in ZEPLIN-II survive anti-coincidence cuts.

Abstract

This article describes the Monte Carlo simulation used to interpret the measurement of the muon-induced neutron flux in the Boulby Underground Laboratory (North Yorkshire, UK), recently performed using a large scintillator veto deployed around the ZEPLIN-II WIMP detector. Version 8.2 of the GEANT4 toolkit was used after relevant benchmarking and validation of neutron production models. In the direct comparison between Monte Carlo and experimental data, we find that the simulation produces a 1.8 times higher neutron rate, which we interpret as over-production in lead by GEANT4. The dominance of this material in neutron production allows us to estimate the absolute neutron yield in lead as (1.31 +/- 0.06) x 10^(-3) neutrons/muon/(g/cm^2) for a mean muon energy of 260 GeV. Simulated nuclear recoils due to muon-induced neutrons in the ZEPLIN-II target volume (~1 year exposure) showed that, although a small rate of events is expected from this source of background in the energy range of interest for dark matter searches, no event survives an anti-coincidence cut with the veto.