Comprehensive study of the background for the Pixel Vertex Detector at Belle II

摘要

The highly successful Belle experiment was located at the KEKB accelerator in Tsukuba, Japan. KEKB was an electron-positron ring accelerator running at the asymmetric energies of 8 GeV (e-) and 3.5 GeV (e+). The Belle experiment took data from 1999 to 2010, but was shut down in June 2010 in order to begin a major upgrade of the accelerator and the detector. Belle played a crucial role in the award of the 2008 Nobel Prize for Physics to M. Kobayashi and T. Maskawa. The main physics goal of Belle was the measurement of CP-violation in the B-meson system.ududThis mission, as well as the search for physics beyond the Standard Model, has been passed to the Belle II experiment located at the SuperKEKB accelerator, the direct successors of the Belle experiment and KEKB respectively. The precise measurement of CP-violation and the search for rare or "forbidden" decays of the B-meson and the tau-lepton as signals for New Physics relies heavily on a large number of recorded events and the precision with which B-meson and lepton decay vertices can be reconstructed. Thus, the accelerator upgrade aims for an increase of the luminosity by a factor of 40, resulting in a peak luminosity of 8x10^35 cm^{-2} s^{-1}. This upgrade is scheduled to be finished by 2017 and will result in asymmetric beam energies of 7 GeV (e-) and 4 GeV (e+), provided by beams with a vertical size of only 48 nm ("nano-beam optics"), a size that has never been reached at any particle collider before.ududThe accelerator upgrade will result in the desired increase of the collision rate of particles, while it will also inevitably lead to an increase in the background for all sub-detectors. The Belle detector would not have been able to handle the new background conditions expected at SuperKEKB, hence an upgrade of the Belle detector to the Belle II detector was necessary. Additionally the upgrade aims to increase the physics performance of the detector, making it more sensitive to the effects of New Physics. The detector upgrade will see improvements and redesigns of almost all subsystems as well as the inclusion of a whole new sub-detector, the PiXel vertex Detector (PXD). The introduction of the PXD will ensure that decay vertices are reconstructed with an extremely high precision in the harsh background conditions expected at Belle II. The PXD is a semi-conductor based particle tracking detector and will be the innermost sub-detector of Belle II. It offers excellent track and vertex reconstruction capabilities, while having a thickness of only 75 μm in order to minimise multiple scattering effects.ududDue to the innovative concept of a high-luminosity nano-beam accelerator, the scale of background being produced at the future SuperKEKB cannot be derived from a traditional electron-positron collider and has, therefore, to be simulated using first-principle Monte Carlo techniques. This thesis focuses on a detailed study of the expected background for the pixel vertex detector at the upcoming Belle II experiment. It starts with a comprehensive summary of the key components of the SuperKEKB accelerator and the Belle II detector before delving into the details of the Belle II simulation and reconstruction framework basf2. It was decided to develop the basf2 framework from scratch, rather than adapting the software framework used at Belle. The changes made in the upgrade from the Belle to the Belle II detector, would have required major modifications of nearly all existing libraries.ududThis thesis continues by explaining, in detail, the measurement and analysis of an experiment conducted at Belle in 2010, shortly before the KEKB accelerator and the Belle detector were shut down. The experiment aimed at establishing the validity of a major background for the PXD, namely the two-photon process into an electron-positron pair, described by the Monte-Carlo generators KoralW and BDK, which have never been tested in the kinematical region relevant for the PXD. From a comparison based on Monte Carlo data it is found that the difference between KoralW and BDK in the high cross-section, low pt region (smaller than 20 MeV) for the produced electron and positron is very small, and that both Monte-Carlo generators agree with the experiment in this important low momentum regime. However, the question arises as to whether the delivered cross-section of the Monte Carlo generators is correct over a wider phase space, but still below the centre-of-mass energies where these generators have been verified experimentally (e.g. at the e+e- colliders PETRA and LEP). In order to answer this question, a comparison between recorded detector data and Monte Carlo data is performed, an analysis that has never been done for centre-of-mass energies of the order of those of the Belle and Belle II experiments. From the results the conclusion is drawn that both Monte Carlo generators, KoralW and BDK, agree very nicely for low values of pt but differ significantly for intermediate values where the total cross-sections are already very small. The recorded data proved that for intermediate pt ranges the behaviour of BDK is correct, while KoralW overshoots the data. Since, however, the cross-section peaks strongly for low values of pt both generators can be used for further background studies.ududFurthermore, this thesis includes a detailed basf2 simulation study of the major beam and QED backgrounds that are expected at Belle II and their impact on the PXD. Various figures of merit are estimated, such as particle flux, radiation dose and occupancy. On average the inner layer experiences a particle flux of 6.1 MHz cm^{-2} and the outer layer of 2.5 MHz cm^{-2}. The distribution of the particle flux along the global z-axis is fairly flat meaning that the radiation damage is evenly distributed along the PXD ladders. The simulation shows that the inner layer of the PXD is exposed to a radiation dose of 19.9 kGy/smy and the outer layer to a dose of 4.9 kGy/smy. Irradiation tests of DEPFET sensors with 10 MeV electrons showed that the sensors work reliably for a dose of at least 100 kGy. It is believed that they can even cope with up to 200 kGy. Using the radiation dose values obtained from the simulation, the numbers translate to a lifetime of roughly 10 years for the PXD sensors, the typical operation time of a high energy physics detector. The study shows that the expected PXD occupancy, summing over all background sources, is given byududinner layer: 1.28 +- 0.03 %udouter layer: 0.45 +- 0.01 %ududThe upper limit for the PXD, imposed by the data acquisition and the track reconstruction, is 3%. The estimated values are well below this limit and, thus, the PXD will withstand the harsh background conditions that are expected at Belle II.