Updated Simulation of GRETA Detector Response and Exploration of Temperature Sensitivity

TL;DR

The paper addresses the fidelity of GRETA/GRETINA signal bases for gamma-ray tracking, focusing on accurate charge transport modeling and electronics response. It introduces a streamlined electronics-response parameterization, reducing free parameters from 669 to 596 and implementing a Python-based superpulse fitting with a trust-region reflective optimizer. By coupling pencil-beam measurements with flood-field experiments and GEANT4-based simulations, it demonstrates that the refined model can compensate mobility changes due to crystal temperature, yielding minimal deterioration in position localization. These findings support using the simplified model in routine GRETA data processing and point to future work on temperature gradients and impurity-profile effects to further improve basis fidelity.

Abstract

The Gamma-Ray Energy Tracking Array (GRETA) is a next-generation gamma-ray spectrometer designed to push the frontiers of nuclear structure and astrophysics experiment. Its high sensitivity is enabled by high-precision localization of gamma-ray interactions within its active detector volume, and the subsequent tracking of gamma-ray scattering sequences. In order to perform gamma-ray tracking, we need to simulate signal generation in the detectors accurately. This requires both accurate calculations of charge movement in the semiconductor volume, as well as a faithful reproduction of real-world experimental effects such as the electronics response. This work addresses the fidelity of the calculated signals for GRETA in two ways. An updated approach has been applied to find an optimized parameterization of the electronics response, while the impact of the detector temperature was also explored to best reproduce experimental signals and improve the position localization performance for GRETA. The results suggest that the electronics response can be simplified without impacting performance, and that the response correction parameters can effectively compensate for the changes in signal which arise due to the crystal temperature, resulting in minimal sensitivity of the position resolution to the assumed temperature.

Figures (8)

• Figure 1: Example of a superpulse (both experimental and simulated) concatenated pulse for events in which segment 0 (first segment on the left) records a net charge. The full superpulse is the set of these concatenated pulses for each segment recording a net charge.
• Figure 2: Localized interaction positions for a typical pencil beam measurement, plotted in a series of 2D projected histograms after rotation of the intrinsic crystal coordinates to account for the tilt of the crystal during measurement. The leftmost figure shows a projection in (x', y'), the center figure in (x', z') and the right-most figure in (y', z'). Brighter regions indicate more interactions occurring at those positions. This figure shows all interaction positions, with no event selection cuts.
• Figure 3: Top panel: x' projection of pencil beam for crystal A35 after signal decomposition using the original crystal basis with 669 electronics response parameters (orange) and the reduced parameter set (blue). Bottom panel: The same as above, but for the y' projection.
• Figure 4: Rising edge of experimental signal (blue) as compared to the pristine signal calculated at a range of crystal temperatures. The sharper rise at lower temperatures is apparent.
• Figure 5: Mean squared error (MSE) plotted versus temperature for the superpulse fitting of crystal A35. The trend is similar for A36, though the measured MSE are slightly lower for that crystal.