WIMPs.

Weakly Interacting Massive Particles (`WIMPs') are the subject of our own experiments as well as those of a number of other groups.

The aim of our collaboration is to detect WIMP dark matter.

Principles of WIMP detection

Collisions between WIMPs and target atoms are little different to billiard ball collisions: the WIMP changes direction, and slows down a little; the atom recoils. This recoil can be detected in four ways:

In semiconductors (such as silicon and germanium) and some liquids and gases, electric charge released by the recoiling atom (`ionization') can be collected and measured electronically.
In certain crystals and liquids (known as `scintillators'), flashes of light are produced as the atom slows down. The amount of light produced depends on the recoil energy.
In crystals, the recoil energy is transformed to vibrations (`phonons'); these can be detected at very low temperatures (close to the absolute zero).
In gases, the recoiling atoms also produce ionization. This can be collected and measured, as in (a), or, in suitable gases, observed as scintillation.

We use technique (b), with the light detected by photomultiplier tubes. Up to 2000, we were mainly using sodium iodide crystals; we now also use liquid xenon, with which it has been possible to obtain much greater sensitivity. By 2003/4, we expect to have installed two-phase (liquid plus gaseous xenon) detectors, using a combination of scintillation and charge collection (a). We are also running a prototype gaseous detector - method (d) - with the goal of observing the directions of the recoiling atoms.

The UK experiments at the Boulby Mine of Cleveland Potash Ltd

Why do astronomy underground?

If WIMPs are the bulk of the galactic dark matter, about a million of them pass through every square centimetre every second, travelling at hundreds of kilometres per second. Shouldn't they be easy to detect?

`Weakly Interacting' means that WIMPs collide very rarely: fewer than one per day will hit an atomic nucleus in a 10 kilogram detector. Meanwhile, we are also being bombarded by cosmic rays - fewer in number, but interacting much more readily. To reduce the unwanted `background noise' from these cosmic rays, our experiments are being carried out in caverns in salt 1100 metres below ground at the bottom of Europe's deepest mine, at Boulby, North Yorkshire. At this depth, collisions in the rock have stopped all but one in a million of the cosmic rays; meanwhile, of the thousand million WIMPs a second passing through you, only about three would collide in the rock on their way down - and they are only slowed down a little, not stopped.

Experiment design

Unfortunately, going deep underground isn't enough to get rid of all the `noise'. Most materials - including the rock walls of the caverns, and the people working there - contain minute traces of natural radioactive atoms, which give out particles whose collisions would also give rise to `noise' in our detectors. These particles are absorbed, as far as is feasible, in high purity shields of either water (by immersing detectors in a tank containing 200 ton(ne)s of pure water) or lead, copper, and wax or polythene. Radioactivity in the detectors themselves must also be minimized, by careful selection of materials (which has involved a lengthy and expensive testing programme). It has not been possible to produce photomultiplier tubes with radioactivity as low as the other detector components, so the scintillator must be shielded by use of silica `light guides' or, as in `NaIAD', by a pure liquid (such as paraffin) which transmits the light flashes from the sodium iodide crystal.

Discriminating between `noise' and `signal'

Despite these efforts, the `signal' rate from dark matter will probably be only one-hundredth to one-thousandth of the radioactivity `noise' rate. So, to obtain the sensitivity we need, we must be able to distinguish the unwanted `background' from events produced by nuclear recoils. In both sodium iodide and liquid xenon (ZEPLIN-I) this is possible because the two kinds of events produce distinguishable light pulses: background events can be rejected by `pulse shape discrimination'. In the two-phase xenon detectors (ZEPLIN-II and III), the relative size of the`primary' (liquid) and `secondary' (gas) signals from background and from nuclear recoils are very different, and should give much improved discrimination. The gaseous ionization detector DRIFT should also give much greater discrimination.