First results from the LUX dark matter experiment at the Sanford Underground Research Facility

TL;DR

The first WIMP search data set is reported, taken during the period from April to August 2013, presenting the analysis of 85.3 live days of data, finding that the LUX data are in disagreement with low-mass W IMP signal interpretations of the results from several recent direct detection experiments.

Abstract

The Large Underground Xenon (LUX) experiment, a dual-phase xenon time-projection chamber operating at the Sanford Underground Research Facility (Lead, South Dakota), was cooled and filled in February 2013. We report results of the first WIMP search dataset, taken during the period April to August 2013, presenting the analysis of 85.3 live-days of data with a fiducial volume of 118 kg. A profile-likelihood analysis technique shows our data to be consistent with the background-only hypothesis, allowing 90% confidence limits to be set on spin-independent WIMP-nucleon elastic scattering with a minimum upper limit on the cross section of $7.6 \times 10^{-46}$ cm$^{2}$ at a WIMP mass of 33 GeV/c$^2$. We find that the LUX data are in strong disagreement with low-mass WIMP signal interpretations of the results from several recent direct detection experiments.

Figures (11)

• Figure 1: Top: Comparison of AmBe data (blue circles) with NEST simulations (blue line), showing excellent agreement above the 2 phe threshold (left axis). The gray histogram and fitted dashed red line show the relative efficiency for detection of nuclear recoils from AmBe data (right axis). Overlaid are the ER detection efficiency from tritium data (green squares), applied to the ER background model in the profile likelihood analysis, and the efficiency from full detector NR simulations treated as real data in terms of the digitized MC-truth S1 phe (purple triangles), applied to the WIMP signal model. The efficiency calculation here does not include S1 or S2 area thresholds. Bottom: NR detection efficiency as a function of nuclear recoil energy for events with a corrected S1 between 2 and 30 phe and a S2 signal greater than 200 phe (black $\bigtriangleup$), the efficiency used directly in the profile likelihood analysis. The efficiency for individually detecting an S2 (red $\Box$) or S1 (blue $\bigcirc$) signal (without the application of any analysis thresholds) are also shown, along with that after the single scatter requirement (green $\bigtriangledown$). The cyan dashed line indicates the threshold in $\textrm{keV}_\textrm{nr}$ below which we assume no light or charge response in the PLR calculation.
• Figure 2: Spatial distribution of all events with position-corrected S1 in the range 2-30 phe from the 85.3 live-days of WIMP search data. The cyan dashed line indicates the fiducial volume. The physical locations of the cathode and gate grids and the detector walls (where the vertical PTFE walls of the TPC form a dodecagon) are also shown.
• Figure 3: Calibrations of detector response in the 118 kg fiducial volume. The ER (tritium, panel $a$) and NR (AmBe and $^{252}$Cf, panel $b$) calibrations are depicted, with the means (solid line) and $\pm 1.28 \sigma$ contours (dashed line). This choice of band width (indicating 10% band tails) is for presentation only. Panel $a$ shows fits to the high statistics tritium data, with fits to simulated NR data shown in panel $b$, representing the parameterizations taken forward to the profile likelihood analysis. The ER plot also shows the NR band mean and vice versa. Gray contours indicate constant energies using an S1--S2 combined energy scale (same contours on each plot). The dot-dashed magenta line delineates the approximate location of the minimum S2 cut.
• Figure 4: The LUX WIMP signal region. Events in the 118 kg fiducial volume during the 85.3 live-day exposure are shown. Lines as shown in Fig. \ref{['ERNRbands']}, with vertical dashed cyan lines showing the 2-30 phe range used for the signal estimation analysis.
• Figure 5: The LUX 90% confidence limit on the spin-independent elastic WIMP-nucleon cross section (blue), together with the ${\pm}1\sigma$ variation from repeated trials, where trials fluctuating below the expected number of events for zero BG are forced to 2.3 (blue shaded). We also show Edelweiss II edelweiss (dark yellow line), CDMS II CDMSLatestResults (green line), ZEPLIN-III ZIII (magenta line), CDMSlite CDMSlite (dark green line), XENON10 S2-only PeterXe10S2only (brown line), SIMPLE SIMPLE (light blue line) and XENON100 100 live-day Xe1002011 (orange line), and 225 live-day Xe100 (red line) results. The inset (same axis units) also shows the regions measured from annual modulation in CoGeNT CoGENTLatestResults (light red, shaded), along with exclusion limits from low threshold re-analysis of CDMS II data CDMS_2keV (upper green line), 95% allowed region from CDMS II silicon detectors CDMS_SiResults (green shaded) and centroid (green x), 90% allowed region from CRESST II CRESST (yellow shaded) and DAMA/LIBRA allowed region DAMA interpreted by DAMA_Savage (grey shaded). Results sourced from DMTools dmtools.