Light detectors for the XENON100 and XENON1T dark matter search experiments

Abstract

One of the most exciting mysteries of modern physics is that of dark matter. While there is a multitude of evidence for its existence, its nature remains unknown. Today, numerous experiments search for weakly interacting massive particles (WIMPs) which could be the constituents of dark matter. Among these is the XENON project. XENON uses a time projection chamber filled with liquid xenon as target, detecting both the prompt scintillation light and the ionisation charge that is created when a particle such as a WIMP scatters off a xenon atom. The XENON100 detector has been taking science data since 2010, leading to the most stringent exclusion limit on spin-independent WIMP-nucleon interactions at the time of writing this thesis, with a minimum of 2 · 10−45 cm2 at 55 GeV/c2 . Its successor, XENON1T is currently being constructed and expected to begin commissioning by the end of 2014. Using 2.2 t of liquid xenon as target it aims on improving upon the sensitivity of XENON100 by two orders of magnitude, with a background expectation of less than one event in a total exposure of 2 t·y. The focus of this thesis lies on the photomultiplier tubes (PMTs) that are used in the XENON100 detector and those that will be used in XENON1T. While XENON100 had already been fully assembled before the start of this PhD project, its 242 PMTs have been subject to a weekly calibration with LED light throughout the entire time the detector was taking science data. This allowed to monitor and ensure the stability of various PMT parameters, namely the gain and the dark count rate. The results of these calibrations and their impact on the science run of 2011/2012 are presented in Chapter 5. For XENON1T a new, larger (3 inch in diameter) type of PMT will be used. To be used in this experiment a PMT must fulfil a multitude of requirements: It must be able to detect single photons with good resolution as well as large light signals of several tens of thousands of photons, which can, during calibration of the detector, occur at a repetition rate of several hundred hertz. At the same time it must be stable over long time periods at the conditions of a cryogenic time projection chamber. It is also necessary that the radioactive contamination of the PMT contributes as little as possible to the overall background of the experiment, and that it does not introduce too much heat into the detector. The PMT has therefore been thoroughly tested, including a long-time stability test in liquid xenon and a test with several cooling cycles. A description of these tests and the results can be found in Chapter 6. The PMT has been found to be suitable for XENON1T and will now be used to instrument the detector. Some of the properties of the PMT behaviour do not only originate from the PMT itself, but also from its voltage divider base. The development of a circuit layout for this base that is optimised for the requirements of XENON1T is described in Chapter 7. The main difficulty was to find a compromise between linearity, which can be improved by increasing the base current and using additional capacitors, power dissipation and the total amount of material, which has to be minimised in order to not introduce additional radioactive background. The selection of the materials that will be used for the bases in XENON1T is also presented. Contributions to the analysis of the electromagnetic background spectrum of the XENON100 experiment are also a part of this thesis. The spectrum has been compared to data from Monte Carlo simulations based on the known radioactivity of the detector materials, showing good agreement. While studying the background, the need for an additional position-dependent correction of the light signals proportional to the ionisation charge was also discovered.