2008 IceCube Update - Section 3
III. EARLY LARGE NEUTRINO DETECTORS
For obvious cost reasons, a 1 km3 detector must use a natural detecting medium. One approach to such a large detector is to search for optical Cherenkov radiation from charged particles produced in neutrino interactions. Three media have been proposed: seawater, freshwater (in a lake), and Antarctic ice. All three have advantages and disadvantages. Water has a very long scattering length but relatively short absorption length. Seawater has high backgrounds from 40K decays and bioluminesence, while the available freshwater lakes suffer from limited size. On the other hand, in ice, the scattering length is shorter than in water, and, once deployed, detector hardware is not recoverable. All three approaches have been pursued. The DUMAND collaboration proposed a large seawater detector back in the 1980's. Currently, the ANTARES, NESTOR and NEMO collaborations are working on detectors in the Mediterranean Sea. A Russian-German collaboration has built a detector in Lake Baikal .
Neutrino detection in ice was pioneered by the AMANDA collaboration. It requires a thick ice sheet, so AMANDA was built at the Amundsen-Scott South Pole station, where the ice is about 2800 m deep. The collaboration drilled holes in the ice using a hot water drill, and lowered strings of optical sensors before the water in the hole refroze.
AMANDA deployed its first string on Christmas Eve 1993, at a depth of 800-1,000 m. It was quickly found that the ice had a very short scattering length, less than 50 cm. This was explained by small (50 μm) air bubbles in the ice. Fortunately, at the higher pressures present at ice depths greater than 1400 m, these bubbles collapse. With this understood, in 1995-6 AMANDA deployed 4 strings with detectors mounted between 1500 and 2000 m deep. These detectors worked as expected, and AMANDA detected its first neutrinos . This success led to AMANDA-II, which, by 2000 consisted of 19 detector strings holding 677 optical sensors. Since 2000, AMANDA-II has been recording about 1,000 neutrino events per year.
However, despite this success, the limitations of AMANDA were becoming obvious. It was too small, and the technology did not lend itself to easy expansion. The AMANDA optical sensors consisted of photomultipliers with resistive bases in a pressure vessel. High voltage was generated on the surface, and analog signals were returned to the surface. Since AMANDA was a prototype detector, several transmission media were tried: coaxial cables, twisted pairs, and, later, optical fibers. The 2.5 km long coaxial cables and twisted pairs dispersed the PMT pulses, with single photoelectron pulses broadened to ∪ μs widths, while the analog optical fibers had a very limited dynamic range. Further, the system was finicky, and not all of the optical modules survived the high pressures present when the water in the drill holes froze. Finally, AMANDA consumed considerable electrical power and required yearly, manpower-intensive calibrations. IceCube was designed to avoid these problems.