Panel Chooses Superconducting Option for the International Linear Collider

The international particle physics community is almost unanimous in its desire for a TeV electron–positron linear collider. Such a facility would be at least 30 km long and cost $5–7 billion. But for more than a decade, competing international collaborations have devoted intensive R&D to two different RF accelerating technologies for the collider (see Physics Today, September 2004, page 49 ). Now, at last, the community has settled on one of the competing technologies.

The International Committee for Future Accelerators has endorsed the recommendation of the 12-member International Technology Recommendation Panel (ITRP) that the RF accelerating cavities be made of superconducting niobium operating at 2 K rather than copper at room temperature (see http://www.ligo.caltech.edu/~skammer/ITRP_Home.htm ). Although the superconducting technology was championed by DESY, the German Electron Synchrotron laboratory in Hamburg, the choice does not imply a specific site for the collider, nor does it imply DESY’s detailed TESLA design. “We’ve chosen a technology, not a specific design,” says ITRP chair Barry Barish of Caltech. The principal laboratories developing the “warm” copper alternative have been SLAC in California and KEK in Japan.

Germany, Japan, and the US are the chief competitors for hosting the collider. It may be another four years before the physicists and governments settle on a site for what is envisioned as a fully international facility—assuming, of course, that the necessary funding is forthcoming.

After seven months of investigation and deliberation, the ITRP concluded that both the cold and warm RF technologies have proven themselves suitable for the demands of a TeV e⁺e⁻ collider. But this “embarrassment of riches,” as Barish calls it, could not continue. “It’s too wasteful to continue pursuing both technologies,” he says. “We can now concentrate the talent, experience, and resources of both teams on the goal of realizing the superconducting machine.”

The ITRP had to weigh a long list of relative strengths and risks. The critical magnetic field intensity above which niobium stops superconducting limits the cold cavity’s accelerating gradient to about two-thirds that of the copper alternative. Thus the cold option requires a total length of 40 km to reach the ultimate goal of two 500-GeV beams colliding head-on. The warm design could do it in 33 km.

On the other hand, the cold option gains several advantages because its RF wavelength, 23 cm, is nine times longer than that of the higher-frequency copper alternative. That makes for larger accelerating cavities less sensitive to small misalignments, seismic ground motion, and the disruptive effects of wake fields on the bunches of beam particles. The copper scheme would require a large number of devices, called SLEDs, to boost accelerating power by temporally compressing the RF pulses coming out of the klystron tubes. A downside of the superconducting scheme’s longer wavelength is that it requires much larger damping rings—that is, storage rings in which synchrotron radiation reduces the phase-space spread of the beams.

A superconducting linac will consume less electric power and require far fewer klystrons than its warm alternative. The klystrons in the warm scheme would also have to be quite close to the accelerating cavities, thus necessitating a second tunnel running alongside the beam tunnel. The superconducting scheme has the option of conveying the RF power to the underground beam tunnel from klystron buildings every few kilometers on the surface. For reasons of safety and access, however, the second tunnel remains under consideration.

An important consideration for the ITRP was that industrialization of the superconducting linac’s major components is already under way. That’s partly because the German government last year approved the construction of the 1.4-km XFEL coherent x-ray light source at DESY (see Physics Today, April 2003, page 35 ). The XFEL will have essentially the same superconducting accelerating structure as the TESLA design. “Aside from the XFEL,” says Barish, “the early involvement of industry in large-scale scientific proposals is more common in Germany than in the US.” Overall, the ITRP report concludes, “The main linac and RF systems [of the superconducting technology] are of comparatively lower risk.”

Early next year, the steering committee for the International Linear collider (ILC), as it will now be called, hopes to appoint a director for the project and set up design and testing teams in Europe, Asia, and the Americas. A site-independent conceptual design should be completed by 2006—after which, work on the detailed engineering design is scheduled to begin, probably even before a site has been chosen.

After final approval by the relevant governments, digging and component manufacturing could begin as early as 2009. By then, first results from the Large Hadron Collider at CERN will have revealed first glimpses of the terra incognita into which both accelerators are to make their complementary forays. For governments not completely convinced by the theoretical arguments that important new physics is bound to show up at LHC and ILC energies, disappointing early LHC results would provide what has been called a last-minute off-ramp.