Electromagnetic calorimeters (EMCs) are one of the key components of present and future hadron and particle physics experiments, since they provide the only possibility for an accurate and efficient energy measurement of electromagnetic probes. The PANDA detector which is planned to be operated at the High Energy Storage Ring (HESR) of the future Facility for Anti-Proton and Ion Research (FAIR) at Darmstadt to study the interaction of cooled anti-protons with a fixed target will apply two different calorimeter concepts. In the target spectrometer, covering the region around the interaction point, a calorimeter based on inorganic scintillation crystals made of lead-tungstate (PbWO4), arranged as a barrel with caps on both sides will be installed to achieve a compact design and an excellent energy resolution. The most forward angles of the setup will be covered by a separate forward spectrometer. In this part, a shashlyk type sampling calorimeter will be implemented. A special advantage for both calorimeters is the coverage of the large dynamic range from 10 MeV up to 15 GeV. Despite the different concepts, the performance of both detectors can be limited by a lateral and longitudinal inhomogeneity of the energy response. This work will discuss the influence of such inhomogeneities on the resolutions of the calorimeters and present concepts to minimize their impact and to improve the performance. For the shashlyk EMC a significant improvement of the lateral homogeneity and a reduction of the light attenuation variation in the modules has been achieved by a new, improved design of the shashlyk modules. With these improvements the lateral inhomogeneity could be reduced from up to ± 17 % to less than ± 2 %. To further optimize the energy resolution, different feature extraction and calibration methods are compared. As a result of all optimizations, a significant improvement of the energy and position resolution has been achieved in the energy range below 1 GeV. The energy resolution for the complete energy range from 50 MeV to 12 GeV can be parametrized by σ/E = 3.54 %/pE/GeV ⊕ 1.34 % and a time resolution close to 100 ps/pE/GeV has been achieved. Altogether it has been proven for the first time that the shashlyk calorimeter of the PANDA detector provides a sufficient performance for the detection of low energetic photons at least down to an energy of 50 MeV. To reproduce the experimental results and to study the impact of different aspects on the energy resolution, the shashlyk EMC prototype has been implemented in GEANT4 and different models for the reproduction of the experimental energy resolution have been compared. For the performance of the barrel part of the target EMC, a special focus has to be set to the energy region below 1 GeV, since most of the detected photons will be in this range. A first crucial parameter for this energy region is the electronic noise. For the current version of the front-end electronics ASIC (APFEL), which is foreseen for the readout of the Avalanche Photo Diodes (APDs) glued to the rear face of the lead tungstate crystals, it has been shown for the first time that a standard deviation of the electronic noise of less than 1 MeV can be achieved. The comparison of different feature extraction methods showed that a fit of the pulse-shape can improve the energy resolution at energies below 100 MeV compared to a simple peak sensing feature extraction. A similar behavior can be observed for an integration of a limited region of the pulse. The main focus of this work is set on the investigation of the influence of the non-uniformity in light collection on the energy resolution. This non-uniformity is introduced by a focusing of the scintillation light due to the tapered side faces of the crystals and causes a position dependence of the energy response, which finally leads to a non-linear response and a deterioration of the energy resolution. It has been shown that a de-polishing of the most tapered lateral side face to an average roughness of 0.3 µm reduces the non-uniformity in light collection of type 2 crystals from originally around 30 % down to a level of less than 5 %. Based on these de-polished crystals, implemented as a 3 × 3 sub-array in the current barrel EMC prototype, the influence of the light collection uniformity on the energy resolution for energies below 1 GeV has been investigated for the first time. It has been proven experimentally that a reduction of the non-uniformity by the de-polishing procedure improves the relative energy resolution for energies above 200 MeV, while the energy resolution between 50 MeV and 200 MeV stays approximately at the same level and only a slight deterioration is expected for energies below 50 MeV. Especially the constant term of the parametrization of the energy resolution has been reduced from more than 2 % to 0.5 % in case of de-polished crystals. To provide a more detailed study of the influence of the non-uniformity on the energy resolution and to obtain a calibration of the prototype with cosmic muons, a detailed model of the current barrel EMC prototype, including the scintillation process and the light collection has been implemented in GEANT4. With a special model, calculating the response of the crystals based on the energy deposition distribution, including the non-uniformity and other empirical values, a good agreement with the experimental results has been achieved.

The PANDA experiment will be one of the flagship experiments at the future Facility for Antiproton and Ion Research (FAIR) in Darmstadt, Germany. It is a versatile detector dedicated to topics in hadron physics, such as charmonium spectroscopy and nucleon structure. A DIRC counter will deliver hadronic particle identification in the barrel part of the PANDA target spectrometer and will cleanly separate kaons with momenta up to 3.5 GeV/c from a large pion background. An alternative DIRC design option, using wide Cherenkov radiator plates instead of narrow bars, would significantly reduce the cost of the system. Compact fused silica photon prisms have many advantages over the traditional stand-off boxes filled with liquid. This work describes the study of these design options, which are important advancements of the DIRC technology in terms of cost and performance. Several new reconstruction methods were developed and will be presented. Prototypes of the DIRC components have been built and tested in particle beam, and the new concepts and approaches were applied. An evaluation of the performance of the designs, feasibility studies with simulations, and a comparison of simulation and prototype tests will be presented.