The protection cases for the couplers and three KEK scientists (left to right: Hiroshi Matsumoto, Kenji Saito,and Sergey Kazakov) in the TOSHIBA clean room in Tochigi, Japan.

The 'warm' side of the coupler.

At the ILC, electrons and positrons will be accelerated in superconducting cavities – tens of thousands of them. The cavities are surrounded by ultra-cold liquid helium (minus 271 degrees Celsius) in cryomodules, giant cold boxes to keep the cavities and liquid helium cold with a temperature difference of nearly 300 degrees Celsius over a distance of 60 cm. Yet each cavity has to be filled with powerful electromagnetic waves to give enough acceleration power to the electrons and positrons.

A klystron generates 500-kilowatt electromagnetic waves at a pulse width of 1.5 milliseconds, five times per second. This momentary energy amount is equal to turning on 1-kilowatt electric heaters in five hundreds houses at the same time. These waves have to be passed through the microwave duct into the cavities without causing significant power loss or wave reflections. This is where a coupler is needed to connect the microwave duct and the cavities in the cryomodules.

When a beam is lost, the energy is reflected backwards. Therefore the electromagnetic waves passing a coupler overlap and correspond to a transfer of power of 2MW, according to a high electric field test. This test proved that a coupler for the cavity can perform a power transfer of 2MW at room temperature. This is one step forward for the ICHIRO-type high electric field cavity, named after the famous Japanese baseball player Ichiro Suzuki for his number, in hope of achieving a high electric field gradient of 51 MV/m.

At KEK, Hiroshi Matsumoto and Sergey Kazakov have been working on couplers for ILC cavities since April 2005. One of the most difficult things in developing couplers is to minimise the heat flowing through a coupler from the outside. In addition to this, it needs to transmit a large amount of power to avoid electric discharge. The two scientists started developing couplers with special consideration of these points.

At the beginning of May, four couplers for ILC cavities were delivered to KEK, and two of them underwent a performance test on 9 June. In this first performance test, they achieved 2.2MW/1.7ms in only 15 hours. This result exceeds the goal of 2MW/1.5ms and meant a huge step forward for high-power couplers. Unfortunately a ceramic window was broken in the end of testing. Technicians brought it into a clean room afterwards and exchanged it with a spare one to perform another high electric field test at the end of June. As a result, they achieved stable running 1MW/1.5ms/5Hz for 15 hours, and 2MW/1.5ms/3Hz for 2.5 hours. In addition, no ceramic windows were damaged this time.

Matsumoto says, "One of our future challenges is to develop strong ceramic." They must not only increase the purity of ceramic, but also improve the technology to meld with metal.

Many researchers were excited by this splendid result. Kazakov, however, restrained his excitement because he had simulated it more than 1,000 times before reaching the current design of the coupler. "My goal is to make the simplest and lowest cost coupler," he says.

-- Nobuko Kobayashi