Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Direct dark matter searches with the full data set of XMASS-I

K. Abe et al. (XMASS Collaboration1)
Phys. Rev. D 108, 083022 – Published 17 October 2023
PDFHTMLExport Citation
Abstract
Authors
Article Text
  • INTRODUCTION
  • DETECTOR
  • CALIBRATION
  • THE DATA SET
  • WIMP SEARCH IN XMASS-I FIDUCIAL VOLUME…
  • ANNUAL MODULATION WIMP SEARCH
  • CONCLUSIONS
  • ACKNOWLEDGMENTS
    • References

    Abstract

    Various weakly interacting massive particle (WIMP) dark matter searches using the full data set of XMASS-I, a single-phase liquid xenon detector, are reported in this paper. Stable XMASS-I data taking accumulated a total live time of 1590.9 days between November 20, 2013 and February 1, 2019 with an analysis threshold of 1.0  keVee. In the latter half of data taking a lower analysis threshold of 0.5  keVee was also available through a new low threshold trigger. Searching for a WIMP signal in the detector’s 97 kg fiducial volume yielded a limit on the WIMP-nucleon scattering cross section of 1.4×10−44  cm2 for a 60  GeV/c2 WIMP at the 90% confidence level. We also searched for WIMP induced annual modulation signatures in the detector’s whole target volume, containing 832 kg of liquid xenon. For nuclear recoils of a 8  GeV/c2 WIMP this analysis yielded a 90% confidence level (CL) cross section limit of 2.3×10−42  cm2. Annual modulation signatures from the Migdal effect and bremsstrahlung at a WIMP mass of 0.5  GeV/c2 are evaluated and lead to 90% CL cross section limits of 1.4×10−35  cm2 and 1.1×10−33  cm2, respectively.

    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    • Figure
    17 More
    • Received 12 November 2022
    • Accepted 13 September 2023

    DOI:https://doi.org/10.1103/PhysRevD.108.083022

    © 2023 American Physical Society

    Physics Subject Headings (PhySH)

    1. Research Areas
    Particle dark matter
    1. Physical Systems
    Weakly interacting massive particles
    1. Techniques
    Dark matter detectors
    Particles & FieldsGravitation, Cosmology & AstrophysicsNuclear Physics

    Authors & Affiliations

    Click to Expand

    Article Text (Subscription Required)

    Click to Expand

    References (Subscription Required)

    Click to Expand
    Issue

    Vol. 108, Iss. 8 — 15 October 2023

    Reuse & Permissions
    Access Options
    Author publication services for translation and copyediting assistance advertisement

    Authorization Required

    Log In
    Other Options
    ×

    Images

    • Figure 1
      Figure 1

      Time evolution of ID SPE gains relative to that in the first week of data taking for each PMT. The black points represent relative gain averaged over all ID PMTs each week. The colored points in the underlying scatter plot show the individual PMTs; the color scale represents the number of PMTs in the 0.0025 gain binning each averaged over that same week. The black dashed vertical line indicates the start of low-threshold data acquisition. The cyan and magenta bands indicate periods when data acquisition was interrupted by the Xe purification work and neutron activation after neutron source calibration.

      Reuse & Permissions
    • Figure 2
      Figure 2

      Time evolution of the weekly dark rate averaged over all PMTs (points), the 98% coverage rate where 98% of the PMTs have a smaller dark rate than this value (dashed), and the highest rate of any single PMT (dotted).

      Reuse & Permissions
    • Figure 3
      Figure 3

      Number of dead ID PMTs over time.

      Reuse & Permissions
    • Figure 4
      Figure 4

      The top panel shows the observed total PE per 122 keV gamma rays, the black markers show Co57 measurements made with the source inside the ID, and the red ones refer to data taken with a Co60 irradiating the LXe target from a fixed position in the OD. The absorption length for LXe scintillation light is shown in the middle panel. The bottom panel shows the scintillation light yield (Ryield) relative to the value at the start, without taking into account the light propagation or PMT effects. The absorption length and the light yield were derived from regular Co57 calibrations.

      Reuse & Permissions
    • Figure 5
      Figure 5

      Relative scintillation efficiency as a function of gamma-ray energy. The efficiency at 122 keV was set to 1. The black data points were measured using gamma-ray sources in the detector, the magenta curve is what we implemented in our MC, and the band represents its one sigma error band. 1.65 keV by Fe55 and 30.0 keV by Am241 are the escape x-ray peaks (escaped into the source container).

      Reuse & Permissions
    • Figure 6
      Figure 6

      Live time accumulation for normal- (blue) and low-threshold (red) data. The black dashed vertical line indicates the start of low-threshold data taking. The magenta bands represent 10-day periods after detector calibration with the Cf252 source. The cyan bands show periods of xenon purification work.

      Reuse & Permissions
    • Figure 7
      Figure 7

      Time evolution of the temperature in and gas pressure above the LXe in the ID. For the temperature, its difference from the nominal temperature (see the text) is shown. The drop at the arrow was due to a change of reference sensor for controlling the cold head that reliquified evaporated Xe from the target mass.

      Reuse & Permissions
    • Figure 8
      Figure 8

      The number of hit distributions for the OD in runs at the beginning (black) and the end (red) of the data-taking period.

      Reuse & Permissions
    • Figure 9
      Figure 9

      Time evolution of the normal threshold event rate after the reduction steps of the standard cut.

      Reuse & Permissions
    • Figure 10
      Figure 10

      Top: PE distributions for each reduction steps after standard cut. Bottom: the reconstructed energy spectrum of data after all reductions. Reconstruction details have been explained in [11, 16, 35].

      Reuse & Permissions
    • Figure 11
      Figure 11

      Energy distributions of the BG simulation after event selection from 0 to 30  keVee. Top: colored stacked histograms showing the contributions from various detector components. Bottom: a cumulative energy spectrum showing the XMASS-I BG-MC with its systematic error band.

      Reuse & Permissions
    • Figure 12
      Figure 12

      Top: the dead PMT correction factor with its associated systematic error. Bottom: the total BG-MC energy spectrum before (dotted) and after (solid) applying the dead PMT correction. The green band shows the range of this correction’s systematic uncertainty.

      Reuse & Permissions
    • Figure 13
      Figure 13

      Data spectra with the statistical error shown in filled dots, and the BG estimate shown in a thick line with the 1σ error from the best fit shown as a shaded band with an energy region between 2  keVee and 30  keVee (top). The WIMP MC expectation for 60  GeV/c2 is also shown in a dotted line with energy region between 2  keVee and 15  keVee. Energy in keVnr is shown on the top. The bottom shows the signal efficiency after applying same event selections.

      Reuse & Permissions
    • Figure 14
      Figure 14

      The spin-independent WIMP-nucleon cross section limit as a function of the WIMP mass at the 90% CL for this work is shown as a solid thick line. Limits [5, 7, 8, 9, 10, 12, 16, 17, 39, 40] as well as allowed regions [41, 42, 43] from other experimental results are also shown.

      Reuse & Permissions
    • Figure 15
      Figure 15

      Time evolution of the event rate before selection (top panel) and after all selections (bottom panel). Blue and magenta show 0.5–20  keVee event rates and 1–20  keVee is shown by black and red.

      Reuse & Permissions
    • Figure 16
      Figure 16

      Energy spectra of observed data before and after event selection, and those of some simulated signal shapes: NR from 8 and 20 GeV WIMPs with 10−41, 10−42  cm2 cross sections, and a Migdal and a bremsstrahlung simulation for 0.5 GeV WIMPs with cross sections of 6×10−34  cm2 and 3×10−32  cm2, respectively.

      Reuse & Permissions
    • Figure 17
      Figure 17

      Left: relative efficiency mean values for both signal (red open circle) and background (black closed circle) events; the overall efficiency was normalized at an absorption length of 8 m for all energy ranges. Right: 1σ ranges of uncertainty in the BG relative efficiencies shown in the left panel. The zero crossing near day 80 is where the improving absorption length passed 8 m.

      Reuse & Permissions
    • Figure 18
      Figure 18

      Dead PMT correction factors for selection efficiency at +1, +4, and +9 dead PMT periods. These corrections are applied to correct the selection efficiency from that derived for the 9 dead PMTs period. After that period, the number of dead PMTs started to increase significantly as shown in Fig. 3. Energy dependence of the correction factor comes from differences in the position distribution of the BG events for each energy region.

      Reuse & Permissions
    • Figure 19
      Figure 19

      The observed event rate with its best fit and expected time variation for an 8  GeV/c2 WIMP signal within the energy ranges of 1.0–1.5, 1.5–2.0, and 19.5–20.0  keVee as examples of the lowest energy bins and the highest energy bin, respectively. The black points indicate data with vertical error bars reflecting the statistical uncertainty of the count rate. The red brackets indicate the 1σ systematic error for each time bin. The green line indicates the best-fit result for the 8  GeV/c2 WIMP spectrum with the decaying BG. All data points and lines are corrected for efficiency scaled with the best-fit’s α.

      Reuse & Permissions
    • Figure 20
      Figure 20

      Limits on the spin-independent elastic WIMP-nucleon cross section as a function of WIMP mass. The solid red line shows the XMASS 90% CL exclusion from the annual modulation analysis. The solid black line shows the 2018 XMASS result [17]. The ±1σ and ±2σ bands represent the expected 90% exclusion regions. Limits, as well as allowed regions from other searches based on event counting, are also shown [5, 7, 41, 42, 43, 47].

      Reuse & Permissions
    • Figure 21
      Figure 21

      Expected energy spectra from Migdal effect and bremsstrahlung caused by 0.4  GeV/c2 WIMPs (magenta for Migdal and cyan for bremsstrahlung) and 1.0  GeV/c2 WIMPs (red for Migdal, blue for bremsstrahlung) with a cross section of 10−35  cm2. The dotted line shows the respective spectra in June and December, while the solid one represents the annual average spectrum. No detector effects are considered yet in these spectra.

      Reuse & Permissions
    • Figure 22
      Figure 22

      Summary of the sub-GeV DM analysis results considering annual modulation with the Migdal effect and the bremsstrahlung. The red solid line is the result of the Migdal analysis for 0.35–4  GeV/c2 WIMP mass, the red dotted line is the result of the bremsstrahlung analysis for 0.32–1  GeV/c2. Limits from other experiments obtained via Migdal and bremsstrahlung effect are also shown for comparison [6, 51, 52, 53].

      Reuse & Permissions
    • Figure 23
      Figure 23

      Modulation amplitude as a function of energy for the model-independent analysis using the pull method (black cross). Solid lines represent the 90% positive (negative) upper limits on the amplitude. The ±1σ and ±2σ bands represent the expected amplitude fluctuation ranges for a null result (for details, see text). The small figure at bottom right shows the signal efficiency used in the model-independent analysis, the overall efficiency for a uniformly distributed electron signal after applying all selection criteria. The latest DAMA/LIBRA result (square) is shown for reference [54].

      Reuse & Permissions
    • Figure 24
      Figure 24

      Limits obtained from the various analyses presented in this paper. Solid lines are the new results, using the full XMASS-I data sets, presented in this paper, in particular, the fiducial volume analysis, the muliti-GeV modulation analysis via NR, and the sub-GeV modulation analysis via Migdal effect and bremsstrahlung. Dotted lines show our earlier results published in [16, 17, 18].

      Reuse & Permissions
    ×

    Sign up to receive regular email alerts from Physical Review D

    Log In

    Cancel
    ×

    Search

    Article Lookup

    Paste a citation or DOI
    Enter a citation
    ×