[Paper Review] The CRESST-II Experiment

R. F. Lang|ArXiv.org|May 30, 2008

Dark Matter and Cosmic Phenomena13 references3 citations

TL;DR

The CRESST-II experiment uses low-temperature scintillating CaWO₄ crystals as targets to search for coherent elastic scattering of weakly interacting massive particles (WIMPs) in a laboratory. By measuring both phonon and scintillation light signals, it achieves background discrimination and sets a new limit of <6×10⁻⁷ pb on the spin-independent WIMP-nucleon scattering cross-section for WIMP masses between 40 and 90 GeV/c², based on a 67 kg·d exposure.

ABSTRACT

The CRESST-II experiment is introduced. It uses scintillating crystals as a target to search for elastic scatterings of dark matter particles. The detectors are operated in a dilution cryostat at temperatures below 30mK, and for each particle interaction, the phonon signal as well as the scintillation light signal are recorded. Recent results from a commissioning run following a major upgrade are presented. The current limit that can be placed on the spin-independent WIMP-nucleon scattering cross-section is below 6x10-7 pb for WIMPs in the mass range from about 40 to 90 GeV/c2.

Motivation & Objective

• To search for coherent elastic scattering of dark matter WIMPs in a laboratory environment using cryogenic scintillating crystals.
• To achieve background suppression through simultaneous measurement of phonon and scintillation light signals.
• To set a competitive exclusion limit on the spin-independent WIMP-nucleon scattering cross-section using a low-threshold, ultra-cold detector system.
• To improve sensitivity by minimizing backgrounds via radiopure materials, multi-layer shielding, and a muon veto.

Proposed method

• Cryogenic CaWO₄ crystals are operated at temperatures below 30 mK in a dilution refrigerator to achieve ultra-low thermal noise.
• Tungsten thin-film superconducting transition-edge thermometers detect phonons from nuclear recoils with high energy resolution.
• A second thermometer on a light-absorbing wafer measures scintillation light, enabling discrimination between electron and nuclear recoils.
• Light yield is used as a discrimination metric: electron recoils have a yield of 1, while WIMP-induced tungsten recoils have a yield reduced by a factor of 40.
• The experiment uses a template-based pulse fitting method that allows negative amplitudes to avoid bias from noise.
• Backgrounds are suppressed via multiple shielding layers (lead, copper, polyethylene), radon exclusion via nitrogen flushing, and a muon veto system.

Experimental results

Research questions

• RQ1 What is the sensitivity of a cryogenic scintillating crystal detector to coherent WIMP-nucleus scattering below 40 keV?
• RQ2 Can simultaneous phonon and scintillation light detection effectively suppress electron recoil backgrounds in low-energy WIMP searches?
• RQ3 How do residual backgrounds from neutron interactions and radioactive contaminants affect the signal region in a commissioning run with partial detector operation?
• RQ4 What is the achievable exclusion limit on the spin-independent WIMP-nucleon scattering cross-section with a 67 kg·d exposure using this dual-signal technique?
• RQ5 How does the performance of this detector compare to other direct detection experiments in the same mass range?

Key findings

• The CRESST-II experiment achieved a limit on the spin-independent WIMP-nucleon scattering cross-section of <6×10⁻⁷ pb for WIMP masses between 40 and 90 GeV/c².
• The limit was derived from a total exposure of 67 kg·d, including downtime, using data from two 300 g CaWO₄ crystals during a 2007 commissioning run.
• Four tungsten recoil events were observed in the signal region between 11 keV and 40 keV, consistent with background expectations.
• The light yield of WIMP-induced tungsten recoils is suppressed by a factor of 40 compared to electron recoils, enabling effective discrimination.
• A weak neutron shielding point above the muon veto was identified post-data-taking, suggesting a possible contribution to the observed background in the signal region.
• The exclusion limit from CRESST-II is competitive with other leading experiments such as CDMS, XENON10, EDELWEISS, and KIMS, despite a significantly lower exposure.