Microscopic Simulation of Particle Detectors

Detailed computer simulations are indispensable tools for the development and optimization of modern particle detectors. The interaction of particles with the sensitive medium, giving rise to ionization or excitation of atoms, is stochastic by its nature. The transport of the resulting photons and charge carriers, which eventually generate the observed signal, is also subject to statistical fluctuations. Together with the readout electronics, these processes - which are ultimately governed by the atomic cross-sections for the respective interactions - pose a fundamental limit to the achievable detector performance. Conventional methods for calculating electron drift lines based on macroscopic transport coefficients used to provide an adequate description for traditional gas-based particle detectors such as wire chambers. However, they are not suitable for small-scale devices such as micropattern gas detectors, which have significantly gained importance in recent years. In this thesis, a novel approach, based on semi-classical (``microscopic'') Monte Carlo simulation, is presented. As a first application, the simulation of avalanche fluctuations is discussed. It is shown that the microscopic electron transport method allows, for the first time, a quantitative prediction of gas gain spectra. Further, it is shown that the shape of avalanche size distributions in uniform fields can be understood intuitively in terms of a toy model extracted from the simulation. Stochastic variations in the number of electrons produced along a charged particle track are another determining factor for the resolution and efficiency of a detector. It is shown that the parameters characterizing primary ionization fluctuations, more specifically the so-called W value and the Fano factor, can be calculated accurately using microscopic techniques such that they need no longer be treated as free variables in the simulation. Profiting from recent progress in the determination of Penning transfer probabilities, the influence of excitation transfer on both primary ionization fluctuations and avalanche statistics is examined and a model for the microscopic calculation of Penning effects is proposed. "Garfield'" is a widely used program for the simulation of gas-based particle detectors. In the context of this thesis work, an object-oriented version (Garfield++) of this software package was developed which includes the above-mentioned microscopic methods. The integration of semiconductor detectors in Garfield++, comprising the adaptation of algorithms, modelling of material properties and validation against measurements, constitutes a further topic of the thesis.