The dark at the end of the tunnel? Evidence for the existence of dark matter begins to stack up

FOR more than a decade, physicists running an experiment called DAMA, about 1,400 metres under Gran Sasso, a mountain near L’Aquila in central Italy, have been fighting a lonely battle. They claim to have detected dark matter—an invisible substance thought…

FOR more than a decade, physicists running an experiment called DAMA, about 1,400 metres under Gran Sasso, a mountain near L’Aquila in central Italy, have been fighting a lonely battle. They claim to have detected dark matter—an invisible substance thought to be five times as abundant as the familiar electrons, protons and neutrons of “ordinary” matter—but until a few months ago almost everyone else who runs similar experiments did not believe them. Now, those opponents are being won over. The latest converts are the members of the CRESST collaboration, also based at the Gran Sasso laboratory (the tunnel leading to which is pictured above). They said on September 6th that they, too, had data suggesting the existence of weakly interacting massive particles (WIMPs), the objects of which dark matter is hypothesised to be composed. That followed an announcement in May that an American experiment called CoGeNT had also seen evidence of WIMPs.

Physicists are pretty confident dark matter exists. Astronomical observations first made in the 1930s show there is not nearly enough ordinary matter to hold galaxies together as they spin round. The visible matter of a galaxy, in other words, must be embedded in a halo of something invisible or else it would fly apart. This dark matter is assumed to consist of novel subatomic particles that interact neither electromagnetically (and so are “blind” to light and other kinds of electromagnetic wave) nor via the strong nuclear force that holds atomic nuclei together. Only gravity and a phenomenon called the weak nuclear force, which governs some types of radioactive decay, can bind them—hence the term “weakly interacting”.

To identify dark matter, experiments like DAMA, CRESST and CoGeNT look for weak-force-mediated collisions between atoms on Earth and WIMPs in the dark-matter halo of the Earth’s home galaxy, the Milky Way. Such collisions should cause individual atomic nuclei to recoil, and with the right apparatus such recoils can be observed. To screen out the confounding effects of cosmic rays, though, such experiments are best located underground.

WIMPs grow stronger

DAMA, led by Rita Bernabei of the University of Rome Tor Vergata, measures the recoils in a 250kg crystal of sodium iodide, a material chosen because it can be grown into such large, pure crystals. Over the 15 years of its operation the experiment’s detector has recorded hundreds of thousands of signals. Most are not caused by dark matter. Even burial deep in the Earth does not screen all background signals out. But although Dr Bernabei cannot tell which individual events (if any) are the result of WIMPs, she does see an intriguing pattern. Every year the number of collisions rises, until it peaks late in May. And every year it then falls back to a minimum in late November. Dr Bernabei and her colleagues believe this is strong evidence that some of the recoils are caused by dark matter, because the motion of the Earth round the sun would be added to the speed with which the solar system moves through the halo in the summer (and thus increase the number of WIMPs that sweep through the crystal) and subtracted from it in the winter, making interception of WIMPs respectively easier and more difficult.

Not everyone agrees. So many things vary with the seasons that such an annual cycle might have another explanation. The CRESST collaboration therefore seeks to distinguish the signal from the noise. CRESST’s physicists, led by Franz Pröbst of the Max Planck Institute for Physics in Munich, cool crystals of calcium tungstate (chosen for similar reasons to sodium iodide) down to within a few thousandths of a degree of absolute zero and use special thermometers to measure the tiny temperature increases that occur when subatomic particles collide with the calcium, oxygen or tungsten nuclei and heat the material up. They think they are able to recognise dark-matter collisions by measuring the minuscule flash of light generated in each collision, as well as the heat. Dark-matter collisions would produce less light than non-dark-matter collisions that generated the same amount of heat.

The CRESST collaboration’s recent result, published in arXiv, a research database, is the result of an analysis of nearly two years’ worth of data taken from eight 300-gram calcium-tungstate crystals. This identified 67 collisions that generated the kind of energy expected from a WIMP interaction. Of these, the researchers calculated that fewer than 50 came from the four known types of interference produced by radioactive decay in and around the experimental apparatus. The rest, they say, could therefore be WIMPs.

Statistically, the result appears plausible. The researchers conclude there is less than one chance in 10,000 that all 67 events are caused by known forms of background interference. But that does not necessarily mean they have spotted WIMPs. The accepted threshold for a discovery in particle physics is one chance in 3.5m that the result is an accident, and Dr Pröbst’s colleague Leo Stodolsky points out that the extra collisions could be caused by unknown background sources, or that he and his colleagues may have underestimated the effects of the four known sources.

The CRESST of a wave?

Dan Hooper, a theoretical physicist at Fermilab, in America, is slightly more indulgent. He describes the latest results as a big step forward and believes that CRESST may well be seeing particles of dark matter. He says the results seem to be roughly compatible with the analyses of DAMA and CoGeNT (the latter, like DAMA, having found evidence of a seasonal fluctuation) since the data of all three groups point to the existence of WIMPs with similar masses and interaction strengths.

The results from two other experiments, CDMS (which, like CoGeNT, is located in the Soudan mine in Minnesota) and XENON, a third project at Gran Sasso, do not fit into this neat picture because they indicate interaction strengths significantly below those estimated from the other experiments. Dr Hooper, though, believes that XENON’s researchers may be underestimating the strengths of these interactions. He also says that the CDMS results might be consistent with those of DAMA, CoGeNT and CRESST thanks to uncertainties in both the relative speeds of WIMPs in the galactic halo and the sensitivities of the various detectors.

Narrowing these uncertainties will mean collecting more data. Technicians at CRESST are currently replacing some of its parts in order to reduce contamination levels within the apparatus itself. Dr Stodolsky reckons they should start seeing results from this more advanced set-up within two years. Fortunately, none of the experiments at Gran Sasso succumbed to the L’Aquila earthquake (see article). So, with luck, the moment may soon arrive when physicists can agree that they have at last glimpsed the missing mass of the universe.