 |
 |
||||||||||||||||||||||||
|
Europhysics News (2003) Vol. 34 No. 4 Direct detection of nonbaryonic dark matter Gérard Nollez A dark universe One of the most favoured hypothesis is that this non baryonic dark matter is constituted by a new type, still undiscovered, of elementary weakly interacting massive particle, which could explain the observed structures of our universe1. Generically called WIMPs, acronym for Weakly Interacting Massive Particles, these hypothetical particles would appear as thermal relics from the Big Bang era during which they were created, now trapped in the gravitational potential of galaxies and clusters of galaxies. It appears as a fascinating coincidence that, in order to develop a cosmologically relevant density parameter, of order unity, these WIMPs must have annihilation cross sections typical of the electroweak interactions. Fortunately enough a wealth of WIMP candidates are offered by Supersymmetry extensions of the Standard Model of Particle Physics [3], rather generically present in supergravity and superstring theories, that allow the unification of the four fundamental interactions. A rich spectrum of new elementary particles is predicted by Supersymmetry and in a large class of supersymmetric models, the lightest supersymmetric particle (generally supposed to be the neutralino) becomes stable. Note that this lightest superparticle, not observed at the LEP collider, must necessarily have a mass greater than 40 GeV. On the other hand, in order that its relic density remains consistent with the observed WM , its mass must be < 1 TeV.
Direct detection
principles More quantitatively, for a given density and a given velocity distribution of neutralinos in the galactic halo and in the case of a dominant coupling independent of the nuclear spin, the event rate R is proportional to the product sA2 where s is the elastic scattering cross-section and A the nucleus mass number. The value of the cross section is predicted by the theory, but the values of many theoretical parameters are still unknown and the predictions allow a very broad domain of values: the allowed values of s, normalised to the proton, spread over 7 orders of magnitude, from 10-48 to 10-41 cm2. The corresponding event rate, in a germanium crystal and for a neutralino mass of say 50 GeV varies from 5.10-7 to 5 event per kilogramme of detector and per day. At these ultra-low signal levels the task of the background reduction in the detector is a very demanding challenge: construction of shielding against external radioactivity, drastic selection of ultra-low radioactivity materials, protection against cosmic radiation in underground laboratories. Moreover, the recoil energy induced from galactic neutralinos is low, below 100 keV, and the shape of the recoil energy spectrum is exponentially decreasing. Most of the signal is lost if the detector energy threshold is > 10 keV and the usual shape of any detector background can easily mimic the signal.
A direct signature of the astrophysical origin of a signal is given by its annual modulation: the Sun moves around the galactic centre with a velocity of about 220 km/s, while the Earth revolves around the Sun at a velocity of about 15 km/s that is added in summer and subtracted in winter, resulting in a sinusoidal modulation of the neutralino flux impinging on a terrestrial detector. The resulting event rate modulation is weak, of the order of 5% at most between June and December. Discovering such a weak ripple requires robust statistics on the signal and this type of detection is reserved to very massive detectors of the order of 100 kg or more. Moreover a number of key experimental parameters (energy calibration, detection efficiencies, possible seasonal modulation of the backgrounds, …) must be maintained under control at a level much better than 5 % over very long periods, a very difficult task. To conclude one can try to give a composite picture of the ideal dark matter detector. It is very massive (100 kg or more) to provide evidence of an eventual annual modulation of the signal, but constructed in a segmented form to eliminate multiple interactions which can not arise from neutralinos. Each detector module has a very low nuclear recoil energy threshold (10 keV or lower) and must discriminate, with a very high efficiency, the nuclear (signal) from electron (background) recoils. Ideally various target materials are used to take advantage of the event rate dependence on A2 as an additional signature. The internal and environmental backgrounds must be maintained at very low levels, including the neutron background: a deep underground site insuring the most efficient suppression of cosmic muons is highly desirable, as well as vetoes to tag the few residual muons able to generate neutron showers. And, last but not least, the detector having to continuously work over very long periods of time in a rather inaccessible site, a very reliable technology is required. As one can suspect, this ideal detector does not yet exist, but various groups actively pursue this goal, using various approaches. Experimental
strategies
Efficient nuclear recoil discrimination, on an event by event basis, is only allowed by very recent technologies offering two readout channels for a single event. Cryogenic detectors, working at milli-Kelvin temperatures, have in common a phonon (or heat) channel that measures the energy deposition independently of the nature of the recoiling particle. For cryogenic phonon-ionisation experiments (CDMS [8], EDELWEISS [9]) the second channel measures the deposited charge (different for nuclear or electron recoils), accomplished by the use of semiconductor crystals, Germanium and also Silicon for CDMS (Fig. 2). These experiments have already produced physics results and are among the most sensitive dark matter searches. For cryogenic phonon-light experiments (CRESST [10], ROSEBUD [11] ) the second channel is a scintillation measurement. A large range of scintillating materials is available, an interesting feature for the project of using targets with different mass numbers. First results of CRESST, using 300 g CaWO4modules, are expected in a few months. Also cryogenic but at a much higher temperature (165 K) the ZEPLIN-II experiment will use liquid Xenon as a target. In addition to the scintillation signal in the liquid phase the charges produced in the liquid (more efficiently by electron recoils) are drifted in the surrounding gas phase where they are detected. An experimental device with a target mass up to 30 kg is under construction.
Three detectors are mounted in a dilution cryostat made of selected materials of very low radioactivity and protected from the radioactive environment by various shields. According to Monte Carlo simulations of these shields, based on the measured neutron flux in the experimental hall (1500 neutrons/m2/day for E > 1 MeV), the rate of neutron scattering events above 20 kev is expected to be of the order of 0.04 per kg and per day. This number of nuclear recoils (which mimic true neutralino interactions) determines the ultimate sensitivity of the experiment in the present configuration. Each detector is a 320g cylindrical Ge monocrystal, 70 mm in diameter and 20 mm in height (the most massive Ge cryodetector today), working at a temperature of about 20 mK. A very sensitive thermal sensor (Neutron Transmutation Doped germanium crystal) glued on the target measures the temperature rise induced by a particle interaction (phonon signal). The plane surfaces are metallized for simultaneous charge collection using a bias tension of a few volts (ionisation signal). The top electrode is divided into a central part (defining a fiducial volume) and a guard ring to identify and eliminate events occuring near the detector edges, whose charge collection is rather imperfect (electric field inhomogeneities).
As to the low-background physics data, most of the events exhibit Q-values around 1 (as expected from gamma interactions) while no event is observed in the nuclear recoil band between 20 keV (the experimental threshold) and 100 keV (upper energy relevant for neutralino masses below 10 TeV) for a total exposure based on two different detectors of 11.7 kg.days. At 90% confidence level (CL), the experimental nuclear recoil rate is thus below 0.2 event/kg/day. The results are interpreted in terms of an upper limit (at 90% CL) on the neutralino-proton scattering cross-section shown in Fig 4. Using a somewhat idealised model of the galactic dark matter halo (spherical and isothermal) an event rate can be calculated. Comparison with the measured one gives the limit of the interaction cross-section above which a signal would become visible in the detector. In Fig. 4 the 3s contour corresponding to the event detection in the DAMA experiment is shown: within this contour the blue point marks the central value of that measurement at M = 52 GeV and s = 7.2 10-42 cm2. The EDELWEISS results are incompatible with the existence of such a neutralino. While 9.8 nuclear recoils should have beeen observed between 20 and 64 keV, none are observed. The Poisson probability of such a fluctuation is only 0.006%. Furthermore the EDELWEISS data start to probe some of the supersymmetric models predicting the highest interaction rates. A look at the
future A jump in the mass of the detectors is planned: CRESST-II, about 10 kg, using cryogenic phonon-light CaWO4 modules of 300g; CDMS-II, about 7 kg, using cryogenic phonon-ionisation modules of Ge (250g) and Si (100g); DAMA with a new assembly of NaI scintillating crystals for a total mass of 250 kg; EDELWEISS-II, about 7 kg in a first phase, up to 30 kg in the final version, using cryogenic phonon-ionisation Ge modules of 320 g (Fig. 5); ZEPLIN-II with 30 kg of scintillating liquid Xe (Fig. 6). At present by far the best discriminating detectors, phonon-charge cryogenic detectors will attempt to further improve their rejection performances by identifying surface events, which might be confused with WIMPs by their incomplete charge collection. This is the strategy followed by the CDMS-II and EDELWEISS-II experiments, in proposing to use thin film thermal sensors to identify and reject surface interactions. On the other hand, the background rejection capabilities of the phonon-light rejection scheme will soon be explored by the CRESST-II and ROSEBUD experiments in real low background conditions; similarly, more efficient rejection capabilities of liquid Xe using a two-phase experiment will be tested in the ZEPLIN-II experiment. Finally a thorough understanding and control of the neutron background is required; large area active muon vetoes are therefore being studied, with the objective of limiting the residual neutron rate to about 10-3 evt/kg/day above 10 keV. At the European level, a collaborative effort is at present initiated to design a direct detection experiment at the one-ton scale, that is required to achieve a sensitivity to WIMP interactions at the 10-46 cm2 level. References [2] Krauss L. M., arXiv:astro-ph/0301012 [3] Jungman G., Kamionkowski M. and Griest K., Physics Reports 267, 195, 1996 [4] HDMS : http://www.mpi-hd.mpg.de/non_acc/dm.html [5] IGEX : http://www.unizar.es/lfnae/ipaginas/ip0400.html [6] DAMA: http://www.lngs.infn.it/lngs/htexts/dama/ [7] ZEPLIN: http://hepwww.rl.ac.uk/ukdmc/ukdmc.html [8] CDMS: http://cdms.berkeley.edu/ [9] EDELWEISS: http://edelweiss.in2p3.fr/ [10] CRESST: http://www.lngs.infn.it/site/exppro/cresst/cresst.htm [11] ROSEBUD: http://www.unizar.es/lfnae/ipaginas/ip0400.html [12] Bottino A., Donato F., Fornengo N. and Scopel S., Phys. Rev. D, 63,125003, 2001 [13] Baltz E.A. and Gondolo P., Phys. Rev. Lett., 86, 5004, 2001 1 Neutrinos were once attractive candidates for non-baryonic dark matter.But recent experiments give evidence of a very low neutrino mass: they cannot significantly contribute to WM. Moreover they are “hot” relativistic particles, and “hot”dark matter leads to much more diluted structures in the universe than can be observed. Copyright EPS and EDP Sciences, 2003 |
|||||||||||||||||||||||||
|
|
||||||||||||||||||||||||
0.005).
The two dominant components in our universe, dark energy and dark matter,
are presently not understood and are both of a very mysterious nature!
