UKDMC Home Page
CERN Courier 36 no. 6 (September 1996) pp 17-20
UKDMC Intranet
UKDMC Home Page |
CERN Courier 36 no. 6 (September 1996) pp 17-20 |
UKDMC Intranet |
New results1 from a low background detector in the UK Boulby salt mine represent a significant step forward in experiments seeking to identify the nature of the non-luminous `dark matter' which makes up 90% of the mass of our Galaxy.
This salt mine, in the north of England, is the most recent of the world sites for an increasing number of underground physics experiments, and the first to be funded specifically for dark matter searches. It is the deepest mine in Europe, reducing the cosmic ray flux to a level comparable with the Italian Gran Sasso and Japanese Kamioka sites. It is very much operational, employing over 1000 workers extracting both salt and potash 24 hours a day, and has a single level at 1100m depth with underground roadways extending over an area about 10 km square. The company (Cleveland Potash Ltd) have given generous support to the work, both allowing use of surface facilities and making available three adjacent disused salt caverns, each about 10m x 20m, in which the current and future UK dark matter programme is being housed. These caverns have been dustproofed and equipped with control rooms, electrical supplies, and data links to a surface control room and from there to the participating groups at Rutherford Appleton Laboratory, Imperial College London, and Sheffield University.
The objective of the programme is to search for new heavy weakly interacting particles (WIMPs) which may constitute the majority of the dark matter. Direct gravitational lensing searches for low mass stars (MACHOs) have not seen sufficient events to indicate more than about 20-30% of the dark matter in that form2 leaving a particle explanation still favoured to account for the majority of the non-luminous density (about 0.4 GeV/cm3 at our distance from the Galactic centre). One hypothesis is that the missing particle is the `neutralino' - the lightest particle of supersymmetry theory (typical expected mass 10 - 10000GeV) - which would have been formed in large numbers in the early universe, subsequently clustering in association with normal baryonic matter. Other possibilities are the axion - a boson of mass 10-5 - 10-1 eV, or one of the known neutrinos, for example a tau neutrino of mass of about 30 eV.
Studies of experiments for all types of particle have been in progress in the UK since 1983, with detector R&D since 1987 and funding to establish the Boulby Mine laboratory starting around 1991 as the first UK joint programme between particle physics and astronomy. This underground programme is currently directed specifically towards WIMP searches.
If WIMPs constitute the Galactic dark matter they will be moving with a typical velocity 0.001 of that of light and could be observed by collisions with ordinary nuclei, imparting recoil energies in the typical range 1 - 50 keV. These recoils could be observed by ionization or scintillation detectors, or by low temperature phonon techniques. The problem lies firstly in the very low predicted event rates - typically 0.01 - 0.1 per kg per day for the neutralino, compared with shielded low energy background gamma and beta rates (from target and detector components) 103 - 104 times higher. Secondly, any underground neutrons (produced by uranium and thorium in the rock and by residual muons) can produce nuclear recoils indistinguishable from dark matter interactions.
The Boulby mine shielding systems are of two types. One (Figure 1) is a 6m tank of purified water in which experiments can be suspended and which absorbs both gammas and neutrons from the cavern walls without adding activity of its own. The other type consists of `castles' of low-activity lead and copper, which surround specific detectors and are further surrounded by wax or polythene neutron shielding. However even the highest purity detector systems have too much intrinsic radioactivity to observe the very low predicted event rates, and methods must be found of distinguishing the nuclear recoil events from background.
In a low activity, low threshold, 6 kg sodium iodide detector running at Boulby since 1994, background rejection factors of 10 - 40 have already been achieved by means of pulse time constant discrimination. Although this technique has been known for 40 years as a method of distinguishing alphas from gammas in sodium iodide in the MeV range, below 20 keV it is more difficult since there are fewer photoelectrons to define the pulse shapes and statistical methods are necessary to search for a population of nuclear recoil events within a larger gamma background.
Figure 2 shows the time constant distribution for 6 months of background events together with the distributions obtained from gamma and neutron calibrations. The background pulses are consistent with zero nuclear recoil signal at all energies and new 90% confidence limits have been obtained on possible dark matter signals, arising from either spin-dependent or spin-independent interactions1.
The current limits are now below 10 events/kg/day, and improvements in light collection and energy threshold will reduce this to below 1 event/kg/day. To reach the favoured neutralino rates 0.01 - 0.1 event/kg/day it is proposed to construct a detector based on a liquid xenon target, for which a combination of scintillation and ionisation processes provides a more powerful method of discriminating nuclear recoils from background3.
It is planned that this new experiment, which can subsequently be scaled up to larger masses to reach even lower event rates, will be carried out in collaboration with the liquid xenon group at UCLA, as the first of several future international collaborations under discussion for the Boulby Mine, both for dark matter searches and for new supernova neutrino detectors.