Results from the ZEPLIN-III experiment

Abstract

The majority of matter in the Universe is dark. World wide efforts to understand this dark component of the Universe are underway and the current evidence suggests the existence of a non-relativistic, non-baryonic and weakly interacting massive particle (WIMP). This weakly interacting dark matter should occasionally couple to baryonic matter, primarily through nuclear interactions. The predicted event rates are low (< Ο(1) events/kg/day). The energy deposited is also expected to be low (≲ 50 keV), and so distinguishing a WIMP signal above the radioactive and cosmic backgrounds is a difficult challenge.

The ZEPLIN-III device was designed to meet this challenge and achieve a competitive sensitivity to WIMP-nucleon interactions. The ZEPLIN-III detector is a two phase time projection chamber using liquid xenon as a target. The instrument was designed to detect dark matter by measuring scintillation and ionisation. Measuring two signals produced by incident radiation allows for discrimination between event types. This allows separation of the main component of the background radiation (primarily electron recoils) from any population of WIMP events that may be present in the data. The ZEPLIN-III detector completed its first science run in 2008, achieving a discrimination power of 1:7800 between nuclear and electron recoils, the highest of any liquid xenon detector. This result limited the WIMP-nucleon cross section to less than 8:4 X 10⁻⁸ pb at 90% confidence level (double sided) for a WIMP mass of 55 GeV/c².

The ZEPLIN-III detector then entered an upgrade phase. The two main improvements included the installation of a new ultra-low background PMT array, significantly reducing the main source of background events, and the addition of a veto detector. The veto detector significantly increased the detector's ability to reject WIMP-like background signals, which may be produced by background neutron events. The veto detector also aided background discrimination by detecting 28% of γ-ray events from the fiducial volume of ZEPLIN-III. The second science run of ZEPLIN-III began in June 2010 and continued until May 2011. During the second science run the discrimination power was 1:280 between nuclear and electron recoils. A total of 8 events were observed in the WIMP search region, which is consistent with background expectations. Assuming a null detection allowed the exclusion of the scalar cross-section above 4:8 10⁻⁸ pb near a WIMP mass of 51 GeV/c². This result was combined with the result from a re-analysis of the first science run using more recent results for the relative scintillation yield, Leff, to give a total limit on the spin independent cross-section of 3:9 X 10⁻⁸ pb at 90% confidence near 52 GeV/c² WIMP mass for the ZEPLIN-III experiment. The WIMP-neutron spin-dependent cross-section limit is 8:0 X 10⁻³ pb at 50 GeV/c² at 90% confidence for the combined first and second science runs. At the time of publication, these were the world's second best, and best results, respectively.