After LUX: The LZ Program

TL;DR

The paper addresses the challenge of direct dark matter detection by outlining the LZ program, a two-stage sequence of liquid xenon time projection chamber detectors that build on the LUX experience. It details a comprehensive technology plan spanning PMT radiopurity, cryogenics, materials, electronics, calibrations, scintillator integration, and a large water Cherenkov shield to suppress all major backgrounds, along with extensive background modeling across external, internal, intrinsic, and neutrino sources. The work emphasizes scalable designs and rigorous radiopurity screening to enable a dramatic increase in target mass from the first stage to the final 20 t detector, with sensitivity projections approaching the neutrino floor for spin-independent WIMP interactions. If realized, LZ would push direct detection into a regime where neutrino backgrounds limit discovery potential, representing a significant step toward uncovering the nature of dark matter and guiding future detector concepts.

Abstract

The LZ program consists of two stages of direct dark matter searches using liquid Xe detectors. The first stage will be a 1.5-3 tonne detector, while the last stage will be a 20 tonne detector. Both devices will benefit tremendously from research and development performed for the LUX experiment, a 350 kg liquid Xe dark matter detector currently operating at the Sanford Underground Laboratory. In particular, the technology used for cryogenics and electrical feedthroughs, circulation and purification, low-background materials and shielding techniques, electronics, calibrations, and automated control and recovery systems are all directly scalable from LUX to the LZ detectors. Extensive searches for potential background sources have been performed, with an emphasis on previously undiscovered background sources that may have a significant impact on tonne-scale detectors. The LZ detectors will probe spin-independent interaction cross sections as low as 5E-49 cm2 for 100 GeV WIMPs, which represents the ultimate limit for dark matter detection with liquid xenon technology.

Figures (6)

• Figure 1: Size comparison of the LUX, LZS, and LZD detectors, including LZ scintillator volumes. LZS and LZD represent mass scale-ups beyond LUX by factors of $\times$5-10 and $\times$60, respectively. The modest increase in linear dimension for the LZS active region allows the detector to make use of the existing LUX water tank. LZD will require a 12 m $\times$ 12 m water shield in a separate lab space.
• Figure 2: Single photoelectron and gain measurements with a 7.6 cm Hamamatsu R11065 PMT. The PMT was tested in liquid xenon (-100$^\circ$ C) at 1500 V. These tubes are shown to perform comparably in liquid xenon as compared to the 5.7 cm R8778 PMTs used in LUX.
• Figure 3: Electron drift length in liquid Xe over time in the LUX 0.1 experiment. 1$\sigma$ and 2$\sigma$ error bars are shown for each measurement. 55 kg Xe were circulated and purified while testing final designs for the LUX heat exchanger and weir. The initial purification time constant was measured at 9$\pm$1 hrs, unprecedented for a detector with a thermal mass comparable to LUX 0.1. The system will be tested with LUX during initial running, and is directly scalable for the LZ detectors.
• Figure 4: Simulated nuclear recoil energy spectrum in LZD resulting from high-energy neutrons generated from muon spallation in cavern rock, before (dark) and after (light) application of a scintillator veto cut with a 100 keV threshold. The simulation assumed a 12 m $\times$ 12 m water tank with a 2 m $\times$ 2 m liquid xenon target at its center, surrounded by an additional 1 m thick liquid scintillator region (modeled as water). The initial neutron energy spectrum was obtained from Mei2005. Within simulation statistics, standard fiducialization and single-scatter cuts eliminate all background events at $>$99.9% efficiency. The simulation was performed using the LUXSim package.
• Figure 5: (Left) LZD NR backgrounds from neutrino coherent scattering. The low-energy range (1-50 keV$_{\text{r}}$) where WIMP signatures are most prevalent include coherent scattering contributions from three different neutrino sources: $^8$B solar neutrinos, neutrinos from cosmic ray interactions in the atmosphere, and diffuse supernova background neutrinos. Due to the steepness of the spectrum, the $^8$B solar neutrino contribution is convolved with the expected detector energy resolution in order to estimate leakage into higher energies in the WIMP search energy range. Overlayed for comparison is the predicted neutron background spectrum for both fission and ($\alpha$,n) contributions from the PMTs, assuming the use of R11410 MOD PMTs. WIMP recoil spectra for 100 GeV particles are shown for interaction cross-sections of $10^{-47}$ cm$^2$ and $10^{-48}$ cm$^2$. A NR acceptance of 50% is assumed. (Right) LZD ER backgrounds from p-p solar neutrinos, $^7$Be solar neutrinos, and two-neutrino double-beta decay from $^{136}$Xe, assuming a $2.1\times10^{21}$ yr lifetime as recently reported in Ackerman2011. Overlaid are projected ER background activities from $^{238}$U and $^{232}$Th decays within the PMTs, assuming the use of R11410 MOD PMTs. A 99.5% rejection factor is applied for neutrino and gamma spectra. Overlaid are WIMP signatures converted into the ER energy scale, assuming a conversion of keV$_{\text{r}}$/keV$_{\text{ee}}$=0.3 and a NR acceptance of 50%.