Temperature Dependence of the Electron-Drift Anisotropy and Implications for the Electron-Drift Model

TL;DR

This study investigates how temperature affects electron drift anisotropy in germanium detectors, revealing discrepancies with existing models and proposing modifications to improve simulation accuracy.

Contribution

The paper provides detailed measurements of temperature-dependent electron drift in germanium detectors and introduces a modified drift model to better match experimental data.

Findings

Longitudinal electron drift anisotropy decreases with temperature.

Standard models produce unphysical predictions at higher temperatures.

Modified drift model aligns better with experimental observations.

Abstract

The electron drift in germanium detectors is modeled making many assumptions. Confronted with data, these assumptions have to be revisited. The temperature dependence of the drift of electrons was studied in detail for an n-type segmented point-contact germanium detector. The detector was mounted in a temperature controlled, electrically cooled cryostat. Surface events were induced with collimated 81 keV photons from a $^{133}$Ba source. A detailed analysis of the rise time of pulses collected in surface scans, performed at different temperatures, is presented. The longitudinal anisotropy of the electron drift decreases with rising temperature. A new approach, making use of designated rise-time windows determined by simulations using SolidStateDetectors$.$jl, was used to isolate the longitudinal drift of electrons along different axes to quantify this observation. The measured temperature dependence of the longitudinal drift velocities combined with the standard electron drift model as widely used in relevant simulation packages results in unphysical predictions. A first suggestion to modify the electron-drift model is motivated and described. The results of a first implementation of the modified model in SolidStateDetectors$.$jl are shown. They describe the data reasonably well. A general review of the model and the standard input values for mobilities is suggested.