Saturday, 16 January 2010

Eyes on Xenon

Not the Higgs but dark matter is the true Holy Grail of high energy physics, given that only the purest can hope to discover it. For most of the past decade the leader of the quest to detect the dark matter particle has been the CDMS collaboration. Unfortunately, this amounted to setting better and better limits on the interaction strength of dark matter with nucleons, apart from this shadow of a hint of a possibility of two events announced last year. Although CDMS stays in the game and will continue taking data as super-CDMS, it is bound to lose the yellow shirt soon. For the moment, the primary contender is Xenon100 - a scaled up version of the Xenon10 detector that was in operation in 2006-2007 in Gran Sasso, Italy.

Xenon experiments use a completely different detection technology than solid state detectors such as CDMS. The detector is filled with xenon in the dual liquid/gas phase. When a xenon atom gets hit, it reports this fact to experimenters in two different ways. Photons produced when the atom returns from the excited state is promptly registered by the phototubes located around the detector volume. Besides, the electrons ionized from the atom drift slowly in the applied electric field, and they are registered after some delay. It turns out that the ratio of the scintillation (S1) and the ionization (S2) signals is different for nuclear recoils (that are due to WIMPs, once the experiment is shielded from neutrons) and electron recoils (that are due to ubiquitous backgrounds like photons).

Thus, by measuring the S1/S2 ratio xenon experiments are able reject most of the background. Furthermore, from the two signals and their relative delay it is possible to reconstruct where in the detector volume the hit occurred. Obviously, background events are more likely to occur near the walls of the tank. Therefore making a larger experiment not only increases the probability of registering a dark matter recoil event , but also decreases the background in the central volume - the property referred to as self-shielding. Add to this the good radioactive purity and relative availability of xenon (you just have to smash a lot of lightbulbs), and you understand why big xenon detectors are taking over the field of direct detection.

The prototype Xenon10 detector was not only a proof-of-principle but also a great success story. For some time, Xenon10 was providing the best constraint on the spin-independent WIMP-nucleon cross-section. In fact, it still sets the best limit for the WIMP masses in the 10-50 GeV range, while for larger masses it was later outraced by CDMS. After so much success, the group decided that things are going too smoothly, and set up a huge pillow fight to ease the tension. As a result, Xenon bifurcated into two rival experiments called Xenon and LUX; the latter was banished from sunny Italy into bottomless pits of South Dakota.

The two groups continued, each on its own, scaling up the same technology, each facing an orthogonal set of problems. Apparently, Xenon was the first to pull together. Last year calibrations were made and the physics run is due any time now. According to the official Xenon propaganda, just 40 live days is enough to push the limit on WIMP-nucleon cross section down to 6x$10^{-9}$ picobarns for a 100 GeV WIMP, almost a factor of 10 better than the current CDMS limit of 4x$10^{-8}$. If either of the two events reported by CDMS is really due to dark matter, by this summer we might have a discovery of the century. If not, the quest will continue, with more and more experiments joining in the race. One-ton monster versions of xenon experiments whose sensitivity should reach $10^{-11}$ picobarns are expected in the second half of this decade.

So tons of excitement ahead. As soon as first rumors appear, you know where to look ;-)


wolfgang said...

I guess the 6x10^-9 figure should be 6x10^-10 ?

Jester said...

No, this one's correct. The CDMS limit had a typo, thanks.

Anonymous said...

great post, many thanks