Xenon10 and Noble Liquid Dark Matter Detectors

1 downloads 0 Views 934KB Size Report
Jul 23, 2007 - Abstract In the field of direct searches for WIMP dark matter, noble liquid .... mary light (S1) and secondary ionization (S2) response from the ...
Journal of Low Temperature Physics manuscript No. (will be inserted by the editor)

S. Fiorucci (for the Xenon Collaboration)

Xenon10 and Noble Liquid Dark Matter Detectors

Received July 23, 2007, Revised September 15, 2007

Keywords Dark Matter, Xenon, Noble Liquids Abstract In the field of direct searches for WIMP dark matter, noble liquid detectors have recently proven an increasingly competitive technology. Although less demanding in terms of cryogenics, they are relatively easily scalable to large target masses and can offer good position reconstruction and background rejection power. Here we illustrate the more recent progress of this technology, through the particular example of the Xenon10 experiment. PACS numbers: 95.35.+d, 29.40.Mc, 95.55.Vj

1 Introduction The past decade has seen the domination of cryodetectors in the field of direct dark matter search, with successful experiments such as CDMS, EDELWEISS or CRESST obtaining world-leading sensitivity results 1,2,3 . Today, noble liquid detectors as an alternative technology look very promising with several international teams working in parallel to prove its efficiency. Among those, we can cite the XENON, ZEPLIN 4 and XMASS 5 collaborations, working with liquid xenon, the WARP 6 , DEAP 7 and ArDM 8 collaborations working with liquid argon and the CLEAN project 9 working with liquid neon. In this article we assume that the reader is reasonably familiar with the dark matter search problem (see for example recent review by Gaitskell 10 ) and focus on the motivation, principles and results behind noble liquid detectors. Department of Physics, Brown University, Providence, RI 02912, USA E-mail: [email protected]

2

2 Noble Liquid detectors 2.1 Why noble liquid detectors? The strength of noble liquid detectors resides in the meeting of several beneficial experimental conditions. Firstly, like cryogenic semiconductor detectors they are able to benefit from two independant measurements of the recoil energy through primary scintillation and ionization (WARP, XENON, ZEPLIN II and III), granting them potentially excellent discrimination against the dominant electron recoil background, typically better than 99 %. In some cases pulse shape analysis on the scintillation signal is also performed (CLEAN, DEAP, WARP, XMASS, ZEPLIN1 ). It has also been shown that high scintillation yields were reachable, leading to competitively low energy thresholds of the order of a few keV recoil energy 11,12 . Secondly, compared to semiconductor crystals this technology is more easily scalable to higher masses (up to the tonne level), which will be needed to scan most of the realistic supersymetric parameter space: the increase in mass simply requires a larger detector, while the required instrumentation scales with the area covered by the PMT arrays. Furthermore, provided they include some kind of position reconstruction method, liquid scintillators actually benefit from a large active volume which provides an efficient, potentially instrumented self-shielding buffer against external background sources. Note however that all noble liquids are not equal in the face of large-scale dark matter search. Neon is relatively cheap at ∼$60/kg and contains no natural radioactive isotopes. It also has a boiling point of 20 K which means that virtually all impurities in the liquid will be frozen out. However it is quite light (A=20) and therefore couples less strongly to a WIMP with a mass of 100 GeV or more. In comparison, Argon is heavier (A=40) and significantly cheaper at ∼$2/kg. However it also naturally contains ∼1 Bq/kg of 39 Ar, which can quickly prove a limitant background at low energy and requires isotope separation to get rid of. This operation can jump the price up to above $1000/kg. Krypton is not used for several reasons, among which its weak scintillation yield and problematic radioactive isotope 85 Kr. Xenon (A=131) is expensive at ∼$800/kg but possesses no long-lived natural radioactive isotopes. It contains 85 Kr but that can be reduced to ppb levels with a simple charcoal getter or through distillation. It also has the highest light yield at ∼46 photons/keV (at zero field), compared to ∼40 ph/keV (Ar) and ∼30 ph/keV (Ne). Lastly, xenon has the added property of naturally containing ∼50% of odd-neutron isotopes 129 Xe and 131 Xe, which can be used to explore spin-dependent WIMP-nucleus interactions. Fig 1 presents the differential and integrated interaction rates for a 100 GeV WIMP with various nuclei, assuming a scalar coupling.

2.2 Principle of operation The current leading noble liquid detector technology is the dual-phase (liquidgas) time-projection chamber (TPC). It is used by the ArDM, WARP, XENON and ZEPLIN II experiments. Although other experiments have been designed to 1

DEAP, XMASS and ZEPLIN do not use a secondary ionization energy measurement.

(dash)Rate>Er [kg/day] (line)dN/dEr [/keVr/kg/day]

3

!1

10

mWIMP=100 GeV, !W!N =8.8×10!44 cm2 !2

10

!3

10

!4

10

!5

10

!6

10

!7

10

0

Xe A=131 Ge A= 73 Ar A= 40 Ne A= 20 20

40 60 Recoil Energy, Er [keVr]

80

100

Fig. 1 (Color online) Differential (solid) and integrated (dash) interaction rates for a 100 GeV WIMP with various nuclei: Xe (red), Ge (blue), Ar (green) and Ne (purple). For an energy threshold < 35 keVr, xenon is the most interesting candidate. Color version in the online proceedings.

rely solely on light pulse-shape discrimination in a single liquid phase, we will focus here on scintillation-ionization TPCs through the example of the XENON10 prototype. An event within the active LXe volume will be characterized by the simultaneous detection of primary scintillation photons and secondary ionization electrons, drifted through the liquid by the application of a ∼0.7 kV/cm electric field and extracted at the liquid-gas interface to produce photons by proportional scintillation (E > 10 kV/cm). Ultimately both primary and secondary signals are recorded by PMTs with a time difference between the two ranging from 0 to 85 µs, corresponding to the electron drift time. High purity of the xenon ensures that electrons have a long enough lifetime, measured in our case at 2.0 ± 0.4 ms. The TPC active volume is defined by a Teflon cylinder of 20 cm inner diameter and 15 cm height (Fig. 2). Teflon is used as an effective UV light reflector 13 and electrical insulator. Four stainless steel (SS) mesh electrodes, two in the liquid and two in the gas, with appropriate bias voltages, define the electric fields to drift ionization electrons in the liquid, extract them from the liquid surface and accelerate them in the gas gap. PMTs are arranged in two arrays for maximum light collection efficiency. The bottom array of 41 PMTs is immersed in the liquid, 1.2 cm below the cathode mesh, to efficiently collect the majority of the direct light which is preferentially reflected downwards at the liquid-gas interface. The top array of 48 PMTs, in the gas, detects the majority of the proportional scintillation light. From the distribution of the PMT hits on the top array, the event location in XY can be reconstructed with a position resolution of a few millimeters. The third coordinate is inferred from the drift time measurement, with better than 1 millimeter resolution.

4

Fig. 2 (Color online) 3D representation of the XENON10 detector (turned 90◦ counterclockwise). Both PMT arrays are visible on top and bottom of the active volume. Bottom PMTs are immersed in the liquid 1.5 cm below the cathode grid, while the liquid-gas interface is located ∼1.2 cm below the top array. Also visible are the various feedthroughs and the Pulse Tube Refrigrator column.

The TPC is enclosed in a SS vessel, insulated by a vacuum cryostat, also made of SS. The total SS mass of 180 kg is the largest amount of material surrounding the 15 kg LXe target. An additional 10 kg of LXe fills the regions outside the active volume. Reliable and stable cryogenics is provided by a pulse tube refrigerator (PTR) 14 with sufficient cooling power to liquefy the xenon and maintain the liquid temperature at -93◦ C with a stability better than 0.005◦ C. At this temperature the Xe vapor pressure is ∼2.3 atm. The Xe gas used for the XENON10 experiment was commercially procured with a guaranteed Kr level below 10 part per billion (ppb), from repeated passages through a dedicated cryogenic distillation column. During operation it is also constantly re-purified by passing through a single high temperature SAES 15 getter in a closed circulation system. The XENON10 detector is surrounded by a shield made of 20 cm-thick polyethylene and 20 cm-thick lead, to reduce background from external neutrons (by a factor ∼ 102 ) and gammarays (by a factor ∼ 105 ). At the Gran Sasso depth of 3150 meters water equivalent, the surface muon flux is reduced by a factor of 106 , such that a muon veto was not necessary for the sensitivity reach of XENON10.

3 Recent results Both the WARP and XENON10 prototypes have reported results this year. In the case of WARP, a remaining low-energy background due to 39 Ar has forced the collaboration to use a relatively high threshold of 55 keV in order to put a limit on the WIMP-nucleon cross-section. With an exposure of 96.5 kg.days, WARP

5

Fig. 3 (Color online) Spin-independent WIMP-nucleon cross-section upper limits (90% C.L.) versus WIMP mass. Shown curves are for the previously best published limit (blue, dash) 20 and the current Xenon10 limit (red, solid). Also shown are the WARP limit (purple, solid), the ZEPLIN II limit (green, solid) and the EDELWIESS limit (dark blue, dash). The shaded areas are the Constrained MSSM models 22 (light blue) and CMSSM with the recent improved Standard Model prediction for the B → Xs γ branching ratio 23 (yellow).

reached a sensitivity of ∼ 10−42 cm2 for a 100 GeV/c2 WIMP 16 . It is expected that a new 50 kg.days data set with improved electronics will remove part or all of this background and lead to a more stringent limit sometime this year. The XENON collaboration reported results 17 in april from the blind analysis of 58.6 live-days of WIMP-search data taken between October 6, 2006 and February 14, 2007. Calibration data taken with a 137 Cs provided ∼ 104 electron recoil events in the energy range of interest (4.5 to 26.9 keV), which is ∼1.3 times more statistics than the wimp-search data analyzed. The detector’s response to nuclear recoils was obtained from 12 hours of irradiation in-situ, using a 200 n/s AmBe source shielded by Pb to reduce the associated gamma flux. Energy calibration was obtained with an external 57 Co gamma ray source and with gamma rays from metastable Xe isotopes produced by neutron activation of a 450 g Xe sample, introduced into the detector after the WIMP search data taking. The primary light (S1) and secondary ionization (S2) response from the 131m Xe 164 keV gamma rays, which interact uniformly within the detector, were used to correct the position dependence of the two signals. With a 99% efficiency at a threshold of 4.5 keV nuclear recoil energy (∼4.4 photoelectrons) 11 , in a fiducial volume of 5.4 kg (8 cm out of 10 in radius, 9 cm out of 15 in Z), ten events were observed in the WIMP search window, mostly concentrated at higher energy. Seven events were expected to statistically leak from the electronic recoil population, but in order to set a conservative limit no background subtraction was considered. Figure 3 shows the 90% C.L. upper limits on WIMP-nucleon cross-sections as a function of WIMP mass, using the ”maximum gap” method developped by S.

6

Yellin 18 and standard assumptions for the galactic halo 19 . It is 8.8 × 10−44 cm2 at a WIMP mass of 100 GeV/c2 , a factor of 2.3 lower than the previously best published limit 20 . 4 Conclusion With several international teams adopting this technology over the past few years, a noble liquid dark matter search experiment has now taken the lead in the field. Excellent background rejection capabilities, low energy threshold, position reconstruction and scalability all make this kind of detectors prime contestants in the race toward an eventual supersymmetric dark matter discovery. Acknowledgements This work is supported by the National Science Foundation under grants No. PHY-03-02646 and PHY-04-00596, and by the Department of Energy under Contract No. DEFG02-91ER40688, the CAREER Grant No. PHY-0542066, the Volkswagen Foundation (Germany) and the FCT Grant No. POCI/FIS/60534/2004 (Portugal).

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

D.S. Akerib et al., Phys. Rev. Lett. 96, 011302 (2006) V. Sanlard et al., Phys. Rev. D 71, 122002 (2005) G. Angloher et al., Astropart. Phys. 23, 325-339 (2005) N.J.T. Smith et al, Proc. 4th Int. Workshop Identification of Dark Matter, p302. Singapore: World Sci (2003) M. Yamashita et al, Astropart. Phys. 20:79 (2003) R. Brunetti et al., New Astron. Rev. 49, 265-269 (2005) M.G. Boulay and 1. Hime, Astropart. Phys. 25, 170-182 (2006) A. Rubbia, J. Phys. Conf. Ser. 39, 129-132 (2006) D.N. McKinsey, K.J. Coakley, Astropart. Phys. 22, 355-368 (2005) R.J. Gaitskell, ”Direct Detection of Dark Matter”, Annu. Rev. Nucl. Part. Sci. 2004, 54:315-59 E. Aprile et al., Phys. Rev. Lett. 97, 081302 (2006) T. Shutt et al., NIM A in print (2007); arXiv:astro-ph/0608137v2 M. Yamashita et al., Nucl. Instr. Meth. A 535, 692 (2004) T. Haruyama et al., Adv. Cryog. Eng. 710, 1459 (2004) SAES Getters: http://www.saesgetters.com P. Benetti et al., Submitted to Astropart. Phys. (2007); arXiv:astroph/0701286 J. Angle et al. (XENON Collaboration), submitted to Phys. Rev. Lett. (2007); arXiv:astro-ph/0706.0039 S. Yellin, Phys. Rev. D, 66, 032005 (2002) J.D. Lewin and P.F. Smith, Astropart. Phys. 6, 87 (1996) D.S. Akerib et al. (CDMS Collaboration), Phys. Rev. Lett. 96, 011302 (2006) A. Bottino et al., Phys. Rev. D 69, 037302 (2004) J. Ellis et al., Phys. Rev. D 71, 095007 (2005) L. Roszkowski, R. Ruiz de Austri and R. Trotta, arXiv:hep-ph/0705.2012