Impact of Geant4's Electromagnetic Physics Constructors on Accuracy and Performance of Simulations for Rare Event Searches

TL;DR

This study evaluates how Geant4 electromagnetic physics constructors influence the total energy deposited by radioactive decays in CaWO4 and Ge targets, covering two thickness geometries and six common contaminants. Using 12 constructors and five production-cut values across 24 test cases, the authors apply a GoF-based compatibility framework (KS, AD, and $\chi^2$) plus contingency tests to quantify systematic differences and identify compatible configurations. They find that the G4EmLivermore constructor yields the highest overall compatibility, while G4EmStandardPhysics_option1 and option2 show notable deviations; thin targets amplify sensitivity to model details and production cuts. Computing performance varies strongly with scattering models and cut values, with single-scattering and hybrid approaches being dramatically slower, and production cuts below 1 reducing speed by about an order of magnitude. The results provide practical guidance for selecting EM physics configurations in rare-event simulations and suggest that validating a single compatible configuration against data could suffice for estimating related systematic uncertainties.

Abstract

A primary objective in contemporary low background physics is the search for rare and novel phenomena beyond the Standard Model of particle physics, e.g. the scattering off of a potential Dark Matter particle or the neutrinoless double beta decay. The success of such searches depends on a reliable background prediction via Monte Carlo simulations. A widely used toolkit to construct these simulations is Geant4, which offers the user a wide choice of how to implement the physics of particle interactions. For example, for electromagnetic interactions, Geant4 provides pre-defined sets of implementations: physics constructors. As decay products of radioactive contaminants contribute to the background mainly via electromagnetic interactions, the physics constructor used in a Geant4 simulation may have an impact on the total energy deposition inside the detector target. To facilitate the selection of physics constructors for simulations of experiments that are using CaWO$_4$ and Ge targets, we quantify their impact on the total energy deposition for several test cases. These cases consist of radioactive contaminants commonly encountered, covering energy depositions via $α$, $β$, and $γ$ particles, as well as two examples for the target thickness: thin and bulky. We also consider the computing performance of the studied physics constructors.

Figures (9)

• Figure 1: Geant4 visualisation of 100.0 decays of ^210Pb ($Q$-value of 63.5) each in CaWO_4 targets with a quadratic cross section of 64 x 64 in the $x$-$y$-plane and for two thicknesses along the $z$-axis: (\ref{['fig:geometry:bulky']}) 64 for the bulky target and (\ref{['fig:geometry:thin']}) 100 for the thin target; the targets are placed in vacuum. Gamma rays are shown in green and electron tracks in red. In case of the bulky target, all decay products are absorbed inside the target; in case of the thin target, some low energy decay products can leave the target. If the decay products get immediate absorbed, their tracks are reduced to dots. Due to technicalities of the visualisation, green dots, representing gamma rays, are usually covered by the red dots, representing electrons, and hence mostly not visible.
• Figure 2: Illustration of the applied statistical analysis. For paired data, only McNemar's test was suitable in stage 2; however, other parts of the workflow were repeated with several statistical tests to avoid systematic biases: For unpaired data both Pearson's $\chi^2$ test of independence and Fisher's exact test was used in stage 2; stage 1 of the workflow was always repeated for $\chi^2$ test, Kolmogorov-Smirnov test, and Anderson-Darling test as goodness-of-fit test. For details, see text.
• Figure 3: Impact of physics constructor and production cut on the comparability between spectra of different physics configurations (red histogram) and the reference spectrum (blue histogram) for the example of total energy deposition by contaminants with large $Q$-values (^208Tl,^210Tl) in a 100-thick CaWO_4 target: (\ref{['fig:208Tl_CaWO4_EP']}, \ref{['fig:208Tl_CaWO4_option2']}) for the same production cut value but different physics constructors; (\ref{['fig:210Tl_CaWO4_option1']}, \ref{['fig:210Tl_CaWO4_option1b']}) for the same physics constructor but different production cut values. $p$-values are given for Anderson-Darling (AD), Kolmogorov-Smirnov (KS), and $\chi^2$ tests.
• Figure 4: Examples of spectra for configurations (red histogram), i.e. pairs of physics constructor and production cut value, that are compatible with the reference spectrum (blue histogram) for various contaminants in thin (100-thick, \ref{['fig:228Ra_Ge_Livermore']}, \ref{['fig:210Pb_Ge_WVI']}) and bulky (64-thick, \ref{['fig:210Tl_Ge_option2']}, \ref{['fig:208Tl_Ge_SS']}) Ge targets. $p$-values are given for Anderson-Darling (AD), Kolmogorov-Smirnov (KS), and $\chi^2$ tests.
• Figure 5: Efficiencies of Geant4 physics configurations, i.e. pairs of physics constructor and production cut value, for different goodness-of-fit tests: (\ref{['fig:eff_AD']}) Anderson-Darling (AD), (\ref{['fig:eff_KS']}) Kolmogorov-Smirnov (KS), and (\ref{['fig:eff_Chi2']}) $\chi^2$. The total efficiencies of physics constructors marginalised over the production cut value is shown in (\ref{['fig:total_eff']}).