Title: WARP: a double phase Argon programme for Dark Matter detection
Abstract: WARP (Wimp ARgon Programme) is a double phase Argon detector for Dark Matter detection under construction at Laboratori Nazionali del Gran Sasso. We present recent results obtained operating deep underground a prototype detector with sensitive mass 2.3 litres. 1. WARP: a double phase argon detector for Dark Matter detection. A double phase Argon detector offers unique sensitivity for the search of dark matter in the form of WIMPs: such detector has the highest discrimination of background events in favour of potential WIMP interactions, which are expected to produce low energy Ar recoils with typical energies of a few tens keV. The basic concept of the detector is the measurement of both the scintillation and the ionization produced by particle interactions inside a liquid argon sensitive volume. Two simultaneous criteria can be applied to select Ar recoils eventually produced by WIMPs: i) Prompt scintillation versus ionization. The prompt scintillation light produced by a particle interacting in the liquid argon phase is detected by PMs. The ionization electrons are extracted from the liquid into the gas and accelerated by an appropriate electric field to produce a proportional (high gain), secondary light pulse seen by the same PMs. The pulse ratio S2/S1 of secondary light S2 (from drift time-delayed ionization) over prompt scintillation light S1 is strongly dependent from columnar recombination of the ionising tracks: therefore nuclear recoils produce typical signals with pulse ratio S2/S1 about 60 times lower than electrons. ii) Pulse shape discrimination of primary scintillation: the primary light is emitted with two components with very large difference in decay times (fast 7 ns, and slow 1.8 μs). The relative amount of the slow component strongly depends from the interacting particle, being around 0.7 for electrons and. 0.1 for heavy charged paricles. The WARP liquid argon detector under construction has a sensitive volume of 100 liters. The goal scintillation yield is of the order of 1 collected photoelectron per keV and the detection threshold for the WIMPs 30 keV. A detailed description of the 100 liters detector can be find in reference [1]. 1 INFN and Dept. of Physics University of Pavia: P. Benetti, E. Calligarich, M. Cambiaghi, C. De Vecchi, R. Dolfini, L. Grandi, A. Menegolli, C. Montanari, M. Prata, A. Rappoldi, G.L. Raselli, M. Roncadelli, M. Rossella, C. Rubbia (Spokesperson), C. Vignoli. INFN and Dept. of Physics University of Napoli “Federico II”: F. Carbonara, A.G. Cocco, G. Fiorillo, G. Mangano, R. Santorelli. INFN Laboratori Nazionali del Gran Sasso and University of L’Aquila: F. Cavanna, N. Ferrari, O. Palamara,. L. Pandola. Princeton University, Physics Department: F. Calaprice, C. Galbiati, Y. Zhao. Institute of Nuclear Physics, Krakow : A. Szelc. Figure 1. Energy spectrum observed with the WARP 2.3 liters prototype in the LNGS underground laboratory inside a 10 cm thick Pb shielding. The overlapped red histogram is the expected (montecarlo-simulated) background by interactions of environmental gamma rays. The residual events below 650 keV are produced by Ar and Kr contaminations inside the liquid Argon. Figure 2. Residual energy spectrum after subtraction of the estimated background from environmental gamma rays. The residual spectrum (upper blue curve) is perfectly fitted by the sum of the beta spectra of Ar (green curve, end-point 565 keV, rate 1.1 Bq/litre) and Kr (red curve, end-point 687 keV, rate 0.5 Bq/litre). The vertical scale is expressed in counts/sec/keV. 2. The WARP 2.3 liters prototype detector In order to perfect the detection method, a 2.3 liters prototype detector is in operation at Laboratori Nazionali del Gran Sasso since February 2005. The detector has been equipped, in subsequent phases, with 2’’ and 3’’ PMs made of low background materials for an onsite detailed study of the backgrounds. The structure is a down-scaled version of the 100 liters detector, with field-shaping electrodes and gas to liquid extraction and acceleration grids. The chamber is filled with ultra-purified argon in order to allow for long drift times of free electrons. Purity is maintained stable by means of continuous argon recirculation. 2.1.1. Study of the β and γ detector backgrounds. The overall background of the 2.3 litres prototype installed underground inside a 10 cm thick Pb shielding has been carefully measured and identified. The total trigger rate above a threshold of 30 keV is about 5 Hz. From a detailed study of the energy spectrum shape (Figure 1) it is shown that about 2 Hz are produced by gamma ray interactions from radioactivity of materials surrounding the sensitive volume; the remaining 3 Hz are produced by the β decays of Kr and Ar dissolved in the liquid argon. In particular, the specific activity of Ar was found to be 1.1 ± 0.4 Bq/litre of liquid Argon, in very good agreement with ref. [2]. We notice that no particular care in the selection of materials was adopted, since in this test phase the background itself helps in the identification of the rejection power. Most of the backgrounds will be strongly reduced in the 100 litres setup. Figure 3. R-like events recorded with the 2.3 liters chamber during 13.4 days of live time in june 2005. The plot shows the primary signal energy (in keV) along the drift time, expressed in μs. The fiducial volume is defined by drift times between 10 and 35 microseconds. Figure 4. Energy distribution of R-like events inside the cathode (upper plot), and inside the fiducial volume (lower plot). The red histogram in the lower plot is the result of a simulation of the expected signal from environmental neutrons in the underground area. 2.1.2. Analysis of Recoil-like events. Data recorded during 13.4 days of live time in a run done in june 2005 have been analyzed looking for recoil-like events by applying the two selection criteria described in section 1. About 6.5 millions events have been processed. The spatial and energy distribution of the 580 selected R-like events (see Figure 3) suggests the following origin for the signals: i) R-like events in the cathode region are mostly induced by decays of Rn daughters. Rn is introduced in the chamber during the filling together with the Ar: being electrically neutral it is uniformly distributed inside the chamber. Daughter nuclei, produced into an ionized state, are drifted to the cathode by the electric field, where they stick. Subsequent decays may end up: (a) with the heavy ion entering the cathode and the α or β travelling in the LAr; (b) with the heavy ion travelling in the LAr and producing the observed R-like signal. The two peaks observed in the energy spectrum (Figure 4, upper plot) are coherently explained by the nuclear recoils from α decays Po Pb (ER=110 keV) and Po Bi (ER=144 keV), assuming a light yield of 0.7 photoelectrons/keV. ii) R-like events inside the fiducial volume are induced by environmental neutrons. Both the event rate and the shape of the energy spectrum (Figure 4, lower plot) are compatible with the expected interactions induced by environmental neutrons inside the underground area (represented by the red histogram). The WARP 2.3 liters chamber in operation at LNGS proofs that the double discrimination technique is effective for separation of recoil events. The first results of the 2.3 liters test (with no neutron shielding) show that the observed background is understood, and that recoil-like signals are compatible with the expected neutron background in the underground area. References [1] WARP proposal, available online at http://warp.pv.infn.it/proposal.pdf [2] H.H. Loosli and H. Oeschger, Earth and Plan. Sci. Lett. 7 (1969) 67