Tokyo

Gone in a flash: Super-K (top) before the accident that left many photomultiplier tubes in pieces (bottom). Credit: ICRR, UNIV. TOKYO

Neutrino physicists around the world have been stunned by an accident at Japan's Super-Kamiokande experiment. The damage will put 'Super-K' out of action for at least a year, delaying several research projects.

Super-K consists of a tank filled with 50,000 tonnes of water, buried deep under Mount Ikena, some 240 kilometres northwest of Tokyo. Neutrinos from the Sun and from supernovae, and those formed when cosmic rays strike the atmosphere, barely interact with matter. But on the rare occasions that they collide with electrons, such as those in water molecules, they emit flashes of blue light known as Cerenkov radiation. These can be detected by sensors called photomultiplier tubes (PMTs).

The walls of Super-K were lined with 11,200 PMTs. Since August, the tank had been drained for maintenance. But on 12 November, as it was being refilled, many of the PMTs imploded. “This has shaken the neutrino physics community,” says Francis Halzen of the University of Wisconsin in Madison.

Researchers familiar with such projects say that the pressure of the water may have burst one of the lower sensors, sending shock waves into neighbouring PMTs and bringing about a chain reaction. About 7,000 of the PMTs shattered, and technicians are now conducting tests to see how many of the remaining PMTs are still functioning. Although the accident also appears to have produced a small crack in the tank's lining, resulting in minor leakage, this is expected to require only minor repairs.

Yoji Totsuka, director of the Kamioka Observatory, which runs Super-K, and head of Tokyo University's Institute for Cosmic Ray Research, is urgently attempting to determine the cause of the disaster. He is also trying to reassure international collaborators. “We will rebuild the detector. There is no question,” reads a message on Super-K's website. The education ministry, which oversees the facility, has promised support, but says it will not ask for special funds. This means that the costs of the repairs — estimated at between US$10 million and US$30 million — will have to be taken from other projects.

The PMTs used at Super-K were made by Hamamatsu Photonics. At more than 50 centimetres across, they are twice the diameter of PMTs at most other neutrino facilities. Totsuka acknowledges that the sensors' large size may have made them more fragile. But he argues that the choice was justified by their improved sensitivity. “The advantage is obvious if you consider the results coming out of the experiment,” Totsuka told Nature.

In 1998, Super-K produced the long-awaited, ground-breaking evidence suggesting that neutrinos have mass. But its work is not finished, and the accident will hamper continuing projects. “We can only hope that they are able to restore the detector soon,” says Art McDonald, director of the Sudbury Neutrino Observatory in Ontario, Canada.

The project that will feel the most immediate impact is 'K2K', which shoots a beam of neutrinos 250 kilometres through the Earth's crust from an accelerator at the High Energy Accelerator Research Organization (KEK) in Tsukuba, north of Tokyo. This Japan–US–Korea collaboration aims to study neutrino 'oscillation', the conversion of neutrinos from one form to another which confirms they have mass.

Neutrinos come in three forms, or 'flavours', and in experimental runs up to April this year, muon neutrinos, the flavour shot from KEK, have been detected by Super-K's sensors 44 times. As this falls short of the 66 expected sightings, the results suggest that the particles are turning into another neutrino variety. “This is an indication, but not conclusive evidence,” admits Koichiro Nishikawa, a spokesman for the K2K collaboration.

The plan was to triple the amount of data by 2005, when the proton synchrotron used in the experiment will be shut down. K2K was scheduled to start taking data again in January 2002. Totsuka hopes to have Super-K back in action by January 2003. If so, it may be possible to increase data collection from the originally planned six months per year to eight or more months to make up for the lost time.

Totsuka is now trying to find a quick and relatively cheap way to get Super-K back online. The damaged PMTs cost $3,000 each when they were installed during 1992–96, and Hamamatsu discontinued their production in 1998. One possibility is to halve the number of PMTs. This would reduce Super-K's sensitivity to solar neutrinos and, to a lesser extent, to atmospheric neutrinos. But those released from KEK are of sufficiently high energy for them still to be readily detected.

Totsuka also hopes that the cheaper, more advanced PMTs being developed for a proposed upgrade known as Hyper-Kamiokande, planned to start within a decade or more, might be introduced earlier at Super-K.

The world's neutrino physicists are hoping that the damage to Super-K can soon be repaired. Its absence from the field will be keenly felt, they say. “I am emotionally attached to the Super-K detector,” says John Bahcall of the Institute for Advanced Study in Princeton, New Jersey. “It has done things that I only previously dreamed about.”

→ http://www-sk.icrr.u-tokyo.ac.jp