Benchmarks and applications of the nuclear deexcitation event generator NucDeEx

TL;DR

The paper tackles the challenge of accurately predicting nuclear deexcitation following neutrino–nucleus interactions to improve neutron multiplicity estimates in detectors. It introduces NucDeEx, a general-purpose, open-source deexcitation event generator with a revised treatment of low-lying discrete states and a Hauser-Feshbach approach for continuum decays, plus a dedicated mode for a high-resolution carbon spectral function. Through extensive benchmarks against other generators and experimental data, the work demonstrates that NucDeEx achieves comparable or superior agreement in gamma-ray branching and neutron-related observables, and it documents practical integrations with Geant4 and NEUT to enhance detector simulations and DSNB background estimations. The findings underscore the importance of inverse kinematics measurements and energy-dependent tuning of level densities to reduce generator dependence and systematic uncertainties in next-generation neutrino experiments.

Abstract

Neutron multiplicity is a key observable in recent neutrino experiments that can enhance the sensitivity of various neutrino physics searches. Nuclear deexcitation plays a significant role in neutron emissions associated with neutrino-nucleus interactions. Therefore, precise prediction of this process is essential. To address this need, a general-purpose nuclear deexcitation event generator \textsc{NucDeEx} was developed and released as an open-source package. The treatment of low-lying discrete excited states was updated to better reproduce experimental data. Benchmarks were conducted using existing nuclear deexcitation event generators and experimental data. Application to other simulators, neutrino event generator \textsc{NEUT} and general particle simulation tool \textsc{Geant4}, are also presented.

Figures (10)

• Figure 1: Algorithm of NucDeEx version 2. The excited state to which the transit occurs is determined according to the branching ratios in Table \ref{['tab:br_p_16O']}. This method is designed to be used with the SF BENHAR1994493PhysRevD.72.053005 or INCL++ PhysRevC.87.014606Leray_2013, which do not have clear peaks corresponding to low-lying discrete excited states.
• Figure 2: Neutron energy spectra from deexcitation of $^{11}\text{B}^*$ (top left), $^{11}\text{C}^{*}$ (top right), $^{15}\text{N}^{*}$ (bottom left), and $^{15}\text{O}^{*}$ (bottom right) predicted by event generators. Excitation energy is selected to 16--35 MeV for $^{11}\text{B}^*$ and $^{11}\text{C}^{*}$, while it is 20--40 MeV for $^{15}\text{N}^{*}$ and $^{15}\text{O}^{*}$. The vertical dashed lines represent the detection energy thresholds in the experiment at RCNP YOSOI2003255Yosoi2004YosM:2003. The numbers shown in each panel represent the emission probability of each event generator: total emission probability and that above the threshold are given in this order. All decays both of single-step and multistep are included.
• Figure 3: Comparison of relative branching ratios of $n$ and $d/\alpha$ for $^{11}$B$^*$ with excitation energies of 16--35 MeV. The magenta histograms show the results using NucDeEx. The green and orange histograms show the results from Ref. PhysRevD.107.072006 and Ref. HU2022137183 using TALYS, respectively. Experimental data PANIN2016204 is shown as blue-hatched histogram. The branching ratios shown here account for only single-step decays.
• Figure 4: Comparison of measured and predicted branching ratios of $n$, $p$, $d$, $t$, and $\alpha$ for $^{11}$B$^*$ with 16--35 MeV excitation energy (left) and for $^{15}$N$^*$ with 20--40 MeV excitation energy (right). The branching ratios of $n$ are scaled by a factor of 1/2. The branching ratio of ABLA's $\alpha$ from $^{11}\text{B}^*$ is also scaed by a factor of 1/2. The magenta histograms show the results using NucDeEx. The green and orange histograms show the results from Ref. PhysRevD.107.072006 and Ref. HU2022137183 using TALYS, respectively. The black histograms denote experimental data from Yosoi et al. with statistical errors, and the authors provide the prediction using CASCADE code written with the blue histograms YOSOI2003255Yosoi2004. The hatched or filled histograms represent the branching ratios for single-step decays, and the open histograms represent those for multistep decays.
• Figure 5: The observed and predicted gamma-ray energy spectra from neutron-oxygen interaction at 30 MeV and 250 MeV of neutron beam energies. The black dots represent the experimental data of E525 10.1093/ptep/ptae159. The red and dark blue lines are Geant4 simulations using INCL++ coupled with G4PreCompoundModel and NucDeEx, respectively. NucDeEx shows better agreement than G4PreCompoundModel in all neutron energies. The figures are from Ref. hino2025simulationmodelinvestigationneutronoxygen, and the chi-squared values are summarized in the paper.