Abstract
The production of neutrons in photon-induced nuclear reactions in the giant-dipole-resonance energy domain remains a topic of high interest for various applications, including the activation and decommissioning of electron accelerator facilities, the detection of illicit materials for homeland security, and the evaluation of neutron dose received by patients during radiotherapy treatments. General-purpose Monte-Carlo (MC) simulation codes for particle transport are intensively used to account for photoneutron production in these applications. However, due to the current scarcity of measured photoneutron energy spectra in the literature, experimental validation of MC-simulated photoneutron energy distributions is not always feasible. Therefore, a critical benchmark among simulation results from various MC codes presently appears as the only option to systematically assess their capabilities in accurately simulating photoneutron production for nuclear reactions of interest. In this work, neutron energy spectra from several targets under irradiation by 20 MeV photons are simulated, employing various state-of-the-art MC codes (FLUKA, Geant4, MCNP6, and PHITS) in their default or generally employed settings. A detailed analysis of the simulated neutron spectra allows one to not only assess the performance of various MC codes in applications such as those mentioned above, but also to partially gauge the incurred systematic uncertainty, and to highlight the present need for more comprehensive evaluated nuclear data in this domain. Ultimately, this work suggests that more prudence is required when using MC codes for applications where photonuclear reactions play a dominant role and where not only the production rate but also the energy spectrum of the emitted neutrons matters.