Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Commissioning Run of the CRESST-II Dark Matter Search

G. Angloher, M. Bauer, I. Bavykina, A. Bento, A. Brown, C. Bucci, C. Ciemniak, C. Coppi, G. Deuter, F. von Feilitzsch, D. Hauff, S. Henry, P. Huff, J. Imber, S. Ingleby, C. Isaila, J. Jochum, M. Kiefer, M. Kimmerle, H. Kraus, J. -C. Lanfranchi, R. F. Lang, B. Majorovits, M. Malek, R. McGowan, V. B. Mikhailik, E. Pantic, F. Petricca, S. Pfister, W. Potzel, F. Proebst, W. Rau, S. Roth, K. Rottler, C. Sailer, K. Schaeffner, J. Schmaler, S. Scholl, W. Seidel, L. Stodolsky, A. J. B. Tolhurst, I. Usherov, W. Westphal

TL;DR

The paper reports the 2007 commissioning run of CRESST-II, detailing upgrades that enable large-scale, cryogenic CaWO4 detectors with a dedicated light detector to discriminate nuclear from electron recoils. The study demonstrates the effectiveness of a 45 cm polyethylene neutron shield and a 20-panel muon veto, achieving roughly a 10-fold improvement in the all-nuclear-recoil background region and validating tungsten-recoil identification via light yield through a neutron test. From 47.9 kg-days of exposure across two 300 g modules, three tungsten-recoil candidates yielded a spin-independent WIMP-nucleon cross-section limit with a minimum of $4.8e-7$ pb at $M_{WIMP} \,\approx\,50$ GeV, illustrating the potential of CRESST-II to reach higher sensitivity in future runs. The results confirm the viability of the light-yield discrimination technique and set the stage for larger-scale, lower-threshold dark matter searches with continued hardware optimization and increased exposure.

Abstract

The CRESST cryogenic direct dark matter search at Gran Sasso, searching for WIMPs via nuclear recoil, has been upgraded to CRESST-II by several changes and improvements.We present the results of a commissioning run carried out in 2007. The basic element of CRESST-II is a detector module consisting of a large (~ 300 g) CaWO_4 crystal and a very sensitive smaller (~ 2 g) light detector to detect the scintillation light from the CaWO_4.Information from light-quenching factor studies allows the definition of a region of the energy-light yield plane which corresponds to tungsten recoils. A neutron test is reported which supports the principle of using the light yield to identify the recoiling nucleus. Data obtained with two detector modules for a total exposure of 48 kg-days are presented. Judging by the rate of events in the "all nuclear recoils" acceptance region the apparatus shows a factor ~ten improvement with respect to previous results, which we attribute principally to the presence of the neutron shield. In the "tungsten recoils" acceptance region three events are found, corresponding to a rate of 0.063 per kg-day. Standard assumptions on the dark matter flux, coherent or spin independent interactions,then yield a limit for WIMP-nucleon scattering of 4.8 \times 10^{-7}pb, at M{WIMP} ~50 GeV.

Commissioning Run of the CRESST-II Dark Matter Search

TL;DR

The paper reports the 2007 commissioning run of CRESST-II, detailing upgrades that enable large-scale, cryogenic CaWO4 detectors with a dedicated light detector to discriminate nuclear from electron recoils. The study demonstrates the effectiveness of a 45 cm polyethylene neutron shield and a 20-panel muon veto, achieving roughly a 10-fold improvement in the all-nuclear-recoil background region and validating tungsten-recoil identification via light yield through a neutron test. From 47.9 kg-days of exposure across two 300 g modules, three tungsten-recoil candidates yielded a spin-independent WIMP-nucleon cross-section limit with a minimum of pb at GeV, illustrating the potential of CRESST-II to reach higher sensitivity in future runs. The results confirm the viability of the light-yield discrimination technique and set the stage for larger-scale, lower-threshold dark matter searches with continued hardware optimization and increased exposure.

Abstract

The CRESST cryogenic direct dark matter search at Gran Sasso, searching for WIMPs via nuclear recoil, has been upgraded to CRESST-II by several changes and improvements.We present the results of a commissioning run carried out in 2007. The basic element of CRESST-II is a detector module consisting of a large (~ 300 g) CaWO_4 crystal and a very sensitive smaller (~ 2 g) light detector to detect the scintillation light from the CaWO_4.Information from light-quenching factor studies allows the definition of a region of the energy-light yield plane which corresponds to tungsten recoils. A neutron test is reported which supports the principle of using the light yield to identify the recoiling nucleus. Data obtained with two detector modules for a total exposure of 48 kg-days are presented. Judging by the rate of events in the "all nuclear recoils" acceptance region the apparatus shows a factor ~ten improvement with respect to previous results, which we attribute principally to the presence of the neutron shield. In the "tungsten recoils" acceptance region three events are found, corresponding to a rate of 0.063 per kg-day. Standard assumptions on the dark matter flux, coherent or spin independent interactions,then yield a limit for WIMP-nucleon scattering of 4.8 \times 10^{-7}pb, at M{WIMP} ~50 GeV.

Paper Structure

This paper contains 21 sections, 9 figures.

Table of Contents

  1. Introduction
  2. Experimental Setup
  3. Detector Support Structure ("Carousel")
  4. Sixty-six channel SQUID readout
  5. Neutron Shield
  6. Calibration Source Lift
  7. Muon Veto
  8. Detectors
  9. Module housing
  10. Thermometers (TES's)
  11. Target Detector
  12. Light Detector
  13. Data Taking and Analysis
  14. Signal Processing and Pulse Fitting
  15. Energy Determination
  16. ...and 6 more sections

Figures (9)

  • Figure 1: $^3$He-$^4$He dilution refrigerator, low background detector environment, and the shielding as upgraded for CRESST-II. The detector carousel (CA) is connected to the mixing chamber of the cryostat (CR) by a long copper cold finger (CF) in order to reduce background originating from the dilution refrigerator. The gas-tight radon box (RB) encloses the low background copper (CU) and low background lead shielding (PB). It is covered by a plastic scintillator muon-veto (MV) and a $45\, \mathrm{cm}$ thick polyethylene neutron moderator (PE). Additional granular PE is placed between the baffles in the upper part of the cryostat to close the line of sight for neutrons coming from above.
  • Figure 2: Detector carousel of ultrapure copper, electropolished to reduce surface contamination. The structure can accommodate 33 detector modules (i.e. $10\, \mathrm{kg}$ of target mass) which can be mounted or dismounted individually. The carousel is mounted at the lower end of the cold finger. The whole structure is cooled to $\sim10\, \mathrm{mK}$.
  • Figure 3: CRESST-II readout circuit. The tungsten film thermometer is in parallel with two reference resistors and the input coil of a DC SQUID. The circuit is biased by a constant current.
  • Figure 4: Schematic arrangement of a CRESST detector module. The module consists of two independent detectors: one with the $\, \mathrm{CaWO_4}$ target crystal, providing a total energy measurement, and one with a silicon-on-sapphire (SOS) wafer for measuring the scintillation light from the target crystal. Both detectors are enclosed in a reflective, scintillating housing.
  • Figure 5: Layout and connection scheme of a light detector used in the commissioning run. Aluminum/tungsten phonon collectors surround the tungsten thermometer or TES. A small portion of the gold thermal link is used as the heater. Connections shown are the aluminum bond wires for the SQUID and heater circuits, and a gold bond wire for thermal coupling, leading to an electrically insulated pad on the detector holder.
  • ...and 4 more figures