[Paper Review] MaGe - a Geant4-based Monte Carlo framework for low-background experiments

Yuen-Dat Chan, J. A. Detwiler+15 more|ArXiv.org|Feb 6, 2008

Dark Matter and Cosmic Phenomena17 references6 citations

TL;DR

MaGe is a Geant4-based Monte Carlo framework developed for low-background, low-energy experiments, particularly for neutrinoless double-beta decay searches using HPGe detectors in the Majorana and Gerda experiments. It enables unified, reusable, and validated simulations of complex background sources—such as muon-induced spallation, neutron interactions, and electromagnetic processes—with key improvements in low-energy physics and neutron simulation accuracy through ongoing collaboration with the Geant4 community.

ABSTRACT

A Monte Carlo framework, MaGe, has been developed based on the Geant4 simulation toolkit. Its purpose is to simulate physics processes in low-energy and low-background radiation detectors, specifically for the Majorana and Gerda $^{76}$Ge neutrinoless double-beta decay experiments. This jointly-developed tool is also used to verify the simulation of physics processes relevant to other low-background experiments in Geant4. The MaGe framework contains simulations of prototype experiments and test stands, and is easily extended to incorporate new geometries and configurations while still using the same verified physics processes, tunings, and code framework. This reduces duplication of efforts and improves the robustness of and confidence in the simulation output.

Motivation & Objective

• To develop a unified, robust, and extensible Monte Carlo simulation framework for low-background, low-energy experiments, particularly for neutrinoless double-beta decay detection.
• To reduce duplication of effort and improve simulation reliability by reusing verified physics processes and code across multiple experiments.
• To support detector design, background modeling, calibration, and sensitivity projections through high-fidelity simulation of complex radiation interactions.
• To enable non-expert users to configure simulations via text file macros without recompiling code, enhancing accessibility and maintainability.
• To validate and improve Geant4's simulation capabilities for low-energy electromagnetic and hadronic processes relevant to underground physics experiments.

Proposed method

• Leverages the Geant4 simulation toolkit as a foundation, extending it with experiment-specific physics lists and tuned parameters for low-energy and low-background conditions.
• Uses text-based macro files for runtime configuration of geometry, physics processes, and I/O interfaces, enabling flexible simulation without code recompilation.
• Integrates with external tools such as event generators, material databases, and waveform generators to enable end-to-end simulation of detector response and pulse-shape analysis.
• Employs a modular, object-oriented C++ architecture to support parallel development and long-term maintenance across the Majorana and Gerda collaborations.
• Validates simulation results against data from test stands, including Clover detectors and segmented HPGe detectors, to ensure accuracy in electromagnetic and neutron interaction modeling.
• Identifies and reports discrepancies in Geant4's neutron and low-energy electromagnetic physics models (e.g., incorrect gamma peak positions, missing meta-stable states and internal conversion electrons) to the Geant4 community for resolution.

Experimental results

Research questions

• RQ1 How can a unified, reusable, and maintainable Monte Carlo framework be developed for low-background, low-energy experiments using Geant4?
• RQ2 To what extent does MaGe accurately simulate electromagnetic interactions, neutron-induced backgrounds, and muon spallation in HPGe detectors?
• RQ3 What are the key discrepancies between MaGe/Geant4 simulations and experimental data for neutron interactions, and how can they be corrected?
• RQ4 How effective is MaGe in supporting detector design, background modeling, and sensitivity projections for neutrinoless double-beta decay experiments?
• RQ5 Can MaGe be extended to other low-background experiments with similar detection and background challenges?

Key findings

• MaGe achieves good agreement with experimental data for electromagnetic interactions, with average discrepancies of (5–10)% across various test stands.
• Geant4 simulations in MaGe were found to underestimate muon-induced neutron production by more than a factor of two in high-Z materials, as observed in the CERN NA55 experiment.
• The simulation overestimates neutron attenuation in concrete compared to measurements from the SLAC electron beam dump experiment, prompting the development of correction methods.
• Several Geant4 bugs were identified and reported: incorrect 2223.0-keV gamma peak position (2224.6 keV in simulation), missing meta-stable nuclear states, and no internal conversion electrons from neutron interactions.
• The MaGe framework successfully simulated and validated spallation neutron production and propagation, demonstrating consistency with other codes like Fluka in benchmark comparisons.
• The framework enables end-to-end simulation including pulse-shape analysis, which is critical for background reduction and systematic uncertainty estimation in low-background experiments.