Results from a Search for Dark Matter in the Complete LUX Exposure

We report constraints on spin-independent weakly interacting massive particle (WIMP)-nucleon scattering using a $3.35\times 10^4\mathrm{kg}$ day exposure of the Large Underground Xenon (LUX) experiment. A dual-phase xenon time projection chamber with 250 kg of active mass is operated at the Sanford Underground Research Facility under Lead, South Dakota (USA). With roughly fourfold improvement in sensitivity for high WIMP masses relative to our previous results, this search yields no evidence of WIMP nuclear recoils. At a WIMP mass of $50\mathrm{GeV}{c}^{-2}$, WIMP-nucleon spin-independent cross sections above $2.2\times {10}^{-46}{\mathrm{cm}}^2$ are excluded at the 90% confidence level. When combined with the previously reported LUX exposure, this exclusion strengthens to $1.1\times {10}^{-46}{\mathrm{cm}}^2$ at $50\mathrm{GeV}{c}^{-2}$.

Figure 1

WS2014–16 data passing all selection criteria. Fiducial events within 1 cm of the radial fiducial volume boundary are indicated as unfilled circles to convey their low WIMP-signal probability relative to background models (in particular, the ${}^{206}\mathrm{Pb}$ wall background). Exposure-weighted average ER and NR bands are indicated in blue and red, respectively (mean, 10%, and 90% contours indicated). Of the 16 models used, the scale of model variation is indicated by showing the extrema boundaries (the upper edge of the highest-S2 model and the lower edge of the lowest-S2 model) as fainter dashed lines for both ER and NR. Gray curves indicate a data selection boundary applied before application of the profile likelihood ratio method. Green curves indicate mean (exposure-weighted) energy contours in the ER interpretation (top labels) and NR interpretation (lower labels), with extrema models dashed.

Figure 2

Efficiencies for NR event detection, estimated using simulation with parameters tuned to calibration data. In descending order of efficiency—red: detection of an S2 (and classification as such by analysis); green: detection of an S1 ($\ge 2$ PMTs detecting photons); blue: detection of both an S1 and an S2; black: detection passing analysis selection criteria. Solid curves indicate exposure-weighted means of the 16 calibrated models. The scale of model variation is illustrated by including the efficiencies of the date and $z$ bins with highest and lowest total efficiency (black dashed curves). Below 1.1 keV nuclear recoil energy, the lowest energy for which light yield was measured in [11], efficiency is conservatively assumed to be zero.

Figure 3

Upper limits on the spin-independent elastic WIMP-nucleon cross section at 90% C.L. The solid gray curves show the exclusion curves from LUX WS2013 (95 live days) [9] and LUX WS2014–16 (332 live days, this work). These two data sets are combined to give the full LUX exclusion curve in solid black (“LUX $\mathrm{WS}2013+\mathrm{WS}2014–16$”). The 1– and $2-\sigma$ ranges of background-only trials for this combined result are shown in green and yellow, respectively; the combined LUX $\mathrm{WS}2013+\mathrm{WS}2014–16$ limit curve is power constrained at the $-1\sigma$ level. Also shown are limits from XENON100 [45] (red), DarkSide-50 [46] (orange), and PandaX-II [47] (purple). The expected spectrum of coherent neutrino-nucleus scattering by ${}^8\mathrm{B}$ solar neutrinos can be fit by a WIMP model as in [48], plotted here as a black dot. Parameters favored by SUSY CMSSM [49] before this result are indicated as dark and light gray ($1-$ and $2-\sigma$) filled regions.