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

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.

💡 Research Summary

The paper investigates how the choice of Geant4 electromagnetic (EM) physics constructors influences both the accuracy of total energy deposition predictions and the computational performance of Monte‑Carlo simulations used in rare‑event searches such as dark‑matter direct detection and neutrinoless double‑beta decay experiments. The authors focus on two widely used detector materials, calcium tungstate (CaWO₄) and germanium (Ge), and construct a matrix of test cases that combines (i) two geometrical thicknesses—a “bulky” 64 mm slab and a “thin” 100 µm slab—(ii) six common radioactive contaminants (low‑Q β emitters ²²⁸Ra and ²¹⁰Pb, high‑Q β emitters ²⁰⁸Tl and ²¹⁰Tl, and α emitters ²³⁴U and ²¹¹Bi), and (iii) the two target materials. This yields 24 distinct scenarios (2 materials × 2 thicknesses × 6 nuclides) that span the full range of electromagnetic interactions (ionisation, bremsstrahlung, photo‑absorption, Compton scattering, atomic relaxation, etc.) relevant for background modelling.

Twelve pre‑defined Geant4 EM physics constructors are examined, ranging from the default G4EmStandardPhysics to specialized low‑energy options such as G4EmLivermorePhysics, G4EmPenelopePhysics, and the highly accurate G4EmStandardPhysics option4. Each constructor implements a different combination of scattering models: single‑scattering (explicit Coulomb scattering), multiple‑scattering (Urban or Goudsmit‑Saunderson models), and hybrid approaches (Wentzel‑VI for large‑angle scattering). To explore the effect of secondary production thresholds, five production‑cut values are applied (200 nm, 1 µm, 1 mm, 1 cm, 10 cm), resulting in 60 distinct physics configurations (12 constructors × 5 cuts).

The authors adopt a two‑stage statistical methodology. In the first stage, three goodness‑of‑fit (GoF) tests—Kolmogorov‑Smirnov, χ², and Anderson‑Darling—compare the energy‑deposition spectra obtained with each configuration against a reference configuration defined as G4EmStandardPhysics option4 with the default 1 mm production cut. A p‑value below 0.05 leads to rejection of the null hypothesis that the two spectra are drawn from the same distribution. In the second stage, categorical analysis aggregates the outcomes of the three GoF tests to assess whether observed differences across configurations are statistically significant overall.

Results show that for the bulky (64 mm) targets, virtually all physics constructors are statistically compatible with the reference, because the full decay Q‑value is absorbed within the material regardless of the scattering model. In contrast, for the thin (100 µm) targets, low‑energy β and α particles can escape, and differences become apparent. Configurations that employ explicit single‑scattering (e.g., G4EmStandardPhysics option1) or the most accurate multiple‑scattering model (Goudsmit‑Saunderson, used in option4) reproduce the reference spectra within statistical uncertainties, while those relying solely on the Urban model (options2 and 3) exhibit deviations up to 5–10 % in the low‑energy tail. High‑energy γ‑ray deposition is largely insensitive to the choice of constructor.

Performance measurements reveal a clear trade‑off. Single‑scattering configurations are the most computationally demanding, especially when combined with very small production cuts (e.g., 200 nm), leading to 2–3× longer CPU times compared with multiple‑scattering options. The Urban‑based constructors (options2 and 3) achieve the best speed‑to‑accuracy ratio for bulk simulations, reducing runtime by roughly 30–40 % while maintaining acceptable agreement for high‑energy backgrounds. Hybrid Wentzel‑VI models (option WVI) occupy an intermediate niche, offering modest speed gains with slightly reduced precision.

Based on these findings, the authors propose practical recommendations. When precise modelling of low‑energy β backgrounds is critical—such as in experiments where surface events dominate the background budget—the most accurate constructor (option4) together with a small production cut (≤ 1 µm) should be used despite the higher computational cost. For large‑scale simulations where high‑energy γ backgrounds dominate and computational resources are limited, the faster Urban‑based constructors (options2 or 3) with a moderate cut (1 mm–1 cm) are preferable. The study underscores that the optimal physics constructor depends on detector geometry, material, and the energy regime of interest, and provides a quantitative basis for informed selection in future rare‑event search simulations.