No abstract

Results are presented from radioactivity screening of two models of photomultiplier tubes designed for use in current and future liquid xenon experiments. The Hamamatsu 5.6 cm diameter R8778 PMT, used in the LUX dark matter experiment, has yielded a positive detection of four common radioactive isotopes: 238 U, 232 Th, 40 K, and 60 Co. Screening of LUX materials has rendered backgrounds from other detector materials subdominant to the R8778 contribution. A prototype Hamamatsu 7.6 cm diameter R11410 MOD PMT has also been screened, with benchmark isotope counts measured at <0.4 238 U / <0.3 232 Th / <8.3 40 K / 2.0±0.2 60 Co mBq/PMT. This represents a large reduction, equal to a change of × 1 24 238 U / × 1 9 232 Th / × 1

11Geant4 is a physics process simulation package developed at CERN, initially for high-2 energy physics simulations [1][2][3]. In the majority of high-energy experiments, the primary 3 particles are generated separate from the active detector elements. This provided a clean 4 distinction in the simulation between the machinery used to generate the beam and the 5 hardware used to measure the beam's effects. 6 Over the years, Geant4 has been expanded to make it more useful for experiments at 7 nuclear energies, including the category of low-background experiments such as neutrino 8 research, searches for neutrinoless double-beta decay, and searches for WIMP Dark Matter. 9 This expanded functionality included additional code to handle electromagnetic interactions 10 down to 250 eV in energy, neutron interactions down to thermal energies, radioactive decays, 11 and event generation from an arbitrary volume rather than a point or a beam. 12Historically, a Geant4 simulation of a low-background experiment would be run, record-13 ing energy depositions only from the active detector components. Inevitably, unexpected 14 phenomena required recording data from passive components as well, to account for all en-15 ergy released in an event. Regardless of the time of an interaction, additional code had to 16 be written into the simulation for every component that recorded data. It was rarely known 17 a priori which parts were altering the observed energy depositions, so more and more com-18 ponents had to be included in the data record, leading to a large proliferation of additional 19 code within the simulation. 20In addition to data recording, low-background experiments must pay special attention to 21 the energy sources of each individual component and material within the detector, support 22 structure, shielding, and environment. These sources include cosmic ray spallation, intrinsic 23 radioactivity, and surface contaminants, and multiple sources are frequently required for 24 a single component. Although educated guesses could be made, it is difficult to know 25 beforehand which sources in which components are the most relevant to the experiment. 26Sources therefore have to be added to more and more components, with additional code 27 required for each combination. 28In the end, it is much easier to simply ensure that all parts have the ability to record 29 data and carry multiple radioactive loads. The code to handle data recording and the code 30 to handle intrinsic radioactivity is largely independent of the part itself. This implies the 31 4 need for a set of classes that provide a consistent approach to both requirements. This paper 1 includes details on such a new set of classes. 2The new features described in this paper is useful across multiple current and future 3 experiments involving nuclear-scale energies and low levels of background activity. They 4 were therefore developed into a generalized code base called LUXSim. These features include 5 creating multiple, simultaneous primary particle types and...