To the general public, the most visible part of the particle physics world is the vast particle accelerator complexes where physicists push the energy frontier, places such as Fermilab and CERN. Despite the importance and expense of these machines, some of the most important particle physics results of the last decade have come from a 50,000 ton tank of water in the mountains of Honshu named Super-Kamiokande. Super-K, as it is commonly called, is the world's largest water Cerenkov detector, which observes fast-moving subatomic particles through the radiation they produce when moving faster than the speed of light in water. Super-K is generally used to detect neutrinos, where the water Cerenkov detector allows particles to be detected right from the point where they are produced by the interaction of a neutrino, and it is in this field of particle physics where its most famous results have occurred.
History of Super-Kamiokande
As can be guessed from the form of the name, Super-K is the successor to an earlier, smaller detector, the KamiokaNDE (sic) detector. KamiokaNDE was originally designed to study a completely different problem: proton decay. Many of the grand unified theories proposed in the early 1980's predicted that the familiar proton, building block of atoms, was not absolutely stable as it is in the Standard Model of particle physics, but rather decays with a very long half-life. This half-life was predicted to be much longer than the age of the universe: about 1030 years, or 1020 times the age of the universe.
Detecting such a slow process would seem to be impossible, given that we don't have 1030 years to wait. However, the sheer prevalence of protons allows us to short-circuit this waiting process. A gram of water contains about 1026 protons, so we would expect one of them to decay every 10,000 years or so. Given the vast quantities of water on the face of the Earth, we can get together enough to see one proton decay a year, or tens, or hundreds, even with a 1030 year lifetime. Thus, KamiokaNDE (Kamioka Nucleon Decay Experiment) was born. 1,000 tons of pure water, containing 1035 protons, was placed in a tank under a mountain in Japan, near the town of Kamioka (now part of Hida city) and instrumented to look for the Cerenkov light of the proton's (much lighter) decay products.
The underground location was chosen so that cosmic rays would not light up the sensitive detector with their near-lightspeed passage, but would instead stop in the dense rock above the detector. This shield, though it would be very effective in stopping muons and electrons from cosmic rays, did not stop the neutrinos produced by cosmic rays striking the atmosphere, known as 'atmospheric neutrinos'. These neutrinos would strike nuclei in the water and produce muons which then in turn produce a detector event, making these neutrinos a significant source of background for the proton decay measurement.
KamiokaNDE didn't see proton decay, and neither have any of the other experiments that have gone looking for it. This is an important measurement in itself, as the lower limit found for the proton lifetime has ruled out many of the grand unified theories that motivated the construction of KamiokaNDE. It was KamiokaNDE's background measurements, though, that motivated the construction of its successor. Its measurement of the atmospheric muon neutrino flux was significantly below predictions, leading to an 'atmospheric neutrino problem' parallel to the more famous Solar Neutrino Problem. KamiokaNDE also contributed to the Solar Neutrino Problem by being the first detector to determine that the presumed 'solar' neutrinos were coming from the direction of the Sun, the previous detectors of solar neutrinos having no sensitivity to the direction of the incoming neutrino. Furthermore, KamiokaNDE pioneered neutrino astronomy by observing the burst of neutrinos from Supernova 1987A, leading to further experiments intended to study supernova neutrinos and try to elucidate the role of neutrinos in supernova explosions.
Super-Kamiokande was designed primarily as an atmospheric neutrino detector, though it doubled as a next-generation proton decay experiment. 50 times the volume of KamiokaNDE, Super-K has a much greater target mass as well as greater sensitivity.
The Super-K Detector
The core of Super-K is a cylindrical acrylic water tank filled with 50,000 tons of purified water. This tank is 41m tall and is 39m in diameter. The inner surfaces of the tank are instrumented with 11,000 sensitive photomultiplier tubes, which detect the Cerenkov light from particles in the detector. The complex is buried under 1 km of rock as shielding from the overwhelming background of cosmic rays.
Most neutrino interactions involve the neutrino transforming through the weak nuclear force into an electron or muon, retaining much of its energy. While electrons and muons are reasonably similar particles, the greatly different mass of the two particles causes them to appear quite different in the Super-K detector. The heavy muons move very close to straight through the detector, which given the cone shape of the Cerenkov radiation makes a sharp, well-defined ring of light on the surface of the detector. Electrons, on the other hand, bounce around in the matter, giving a much fuzzier ring. In this way, Super-K can determine whether the incident neutrino is a muon or electron neutrino, which is an important measurement for finding and characterizing neutrino oscillation.
To fully determine the energy of the particle, it is necessary that it both originate and stop within the volume of the detector. As such, an additional 2,000 phototubes were placed in the detector facing outwards, so as to see the passage of high-energy particles as they enter and leave. This 'outer detector' is a common feature of many neutrino detectors, large and small, allowing much better rejection of background events.
Experimental Achievements of Super-K
Super-K began taking data in 1996, and had its first major results available for the Neutrino98 conference in 1998. These results were the first clear evidence for neutrino oscillations and were based on measurements of atmospheric neutrinos produced at different points on the Earth. Since the neutrinos travelling upwards through the Earth into Super-K have travelled much further since being produced than the neutrinos travelling straight downwards into Super-K, they will have had more time to oscillate and thus are less likely to appear as muon neutrinos. As such, the expectation was that there would be fewer up-going muon neutrinos than down-going muon neutrinos, and this is exactly what Super-K observed.
Super-K continued to record atmospheric neutrino data until disaster struck in November 2001. Approximately six thousand of the photomultiplier tubes imploded in a disastrous chain reaction, with each of the implosions producing a shock wave in the detector water that in turn produced further implosions. The loss of half of the photosensors halved the detector's sensitivity, necessitating the replacement of the broken tubes for resumption of atmospheric and solar neutrino detection.
The remaining unbroken phototubes were distributed evenly over the surface of Super-K, making the Super-Kamiokande II detector. This was primarily for the use of the K2K experiment, which used a beam of high-energy neutrinos produced at the KEK particle accelerator 250 km away in Tsukuba. K2K was the first experiment to use an artificial neutrino beam from a particle accelerator, all previous studies of artificial neutrinos being made using nuclear reactors. Given the known direction, energy, and timing of the K2K neutrinos, the reduced sensitivity of Super-K II was not a major obstacle to the re-observation of neutrino oscillations with the K2K beam. In 2005 and 2006 the broken phototubes were gradually replaced, with the fully-restored detector coming online in June 2006 as Super-Kamiokande III.
The Future of Super-K
Beyond refining its previous measurements, Super-K has a major part to play in the further development of neutrino physics. The T2K experiment is the successor to K2K, planned to begin operation in 2009. A much higher-intensity beam from the new J-PARC accelerator in Tokai will be sent over 295 km to Super-K, with the initial goals of making a precision measurement of the muon neutrino survival probability and a first measurement of electron neutrino appearance in a muon neutrino beam. (Previous experiments have only observed a deficit of muon neutrinos which are assumed to have oscillated mainly into tau neutrinos.)
T2K should occupy Super-K's resources for many years to come, especially if a proposed beam upgrade extends its experimental programme. However, plans for a successor detector, dubbed Hyper-Kamiokande, have been suggested. Hyper-K would consist of two 500,000 ton water Cerenkov detectors, whose improved sensitivity would improve both atmospheric neutrino and T2K results, at the same time as the increased mass increased sensitivity to proton decay. Not only are two half-megaton detectors considered more economical than a single monolithic megaton detector, but it allows the second Hyper-K detector to be placed in a different location, 300 km further from Tokai along the T2K beam, in South Korea. This would not affect the atmospheric neutrino and proton decay experiments, but would provide T2K with a greater handle on neutrino oscillations.
Even if Super-K is eventually superseded by Hyper-K, its location and enclosure should prove useful to future experiments; the pioneering KamLAND reactor neutrino experiment located its rather different detector in the original KamiokaNDE chamber, for example. Barring another catastrophe, the particle physics world should continue to see interesting results from Super-K for many years to come.
Sources include my experience as a graduate student on the T2K experiment and the lectures at the 2007 Fermilab Neutrino Physics Summer School (slides archived at http://nuss.fnal.gov/ .