Distributed-coupling Linear Particle Accelerators

Book excerpt: Linear particle accelerators (linacs) are essential for future discovery machines as well as many advanced medical and industrial applications. A linac is formed from a set of cascaded RF cavities (cells). For a typical electron linac, such as the SLAC linear accelerator, RF power is fed to the linac from one point and flows to adjacent cells through the beam tunnel. Consequently, the linac design process requires careful consideration of the coupling between adjacent cells. This limits the ability of the designer to optimize the cell shape for high RF-to-beam efficiency and/or craft the field on the surface for high-gradient operation. We introduce a novel particle accelerator technology that utilizes a periodic feeding network to feed every accelerating cell independently. This eliminates the need for the coupling between cells, giving considerable optimization flexibility for the shape of the accelerator cells. This dissertation discusses the concept behind this topology and presents how such a concept is developed and implemented through a set of key research milestones. The theory of the distributed-coupling linac is presented alongside the associated optimization techniques that take full benefit of the resultant design flexibility. Compared to a conventional linac, our designed and tested structures provide approximately double the shunt impedance. A novel manufacturing technique is enabled by observing that both the cells and the feeding network have planes with no currents passing through them. This allowed the manufacturing of the structure from two blocks. From an economical point of view, this reduces the part count by about two orders of magnitude in comparison to traditional ways of building the structures from half-cell cups. Additionally, this method allows us to assemble the structure without the necessary brazing steps typically needed for traditional linacs. Hence, the copper or doped-copper material hardness properties can be maintained, further enhancing the ability of the surface to resist damage due to cyclic fatigue. Cryogenic operation of normal-conducting linacs substantially reduces their surface resistance and hence improves RF-to-beam efficiency. The reduced losses also reduce the transient temperature rise on the surface, which is the root cause of the surface cyclic fatigue that leads to surface distortions and consequently breakdown events. That cyclic fatigue is further reduced because the copper yield strength is increased at lower temperatures. In this work, we present the first demonstration of high-gradient acceleration of an electron-beam at a cryogenic temperature of 77 K. Experimental operation of the distributed-coupling structure at 77 K resulted in a reduction in the breakdown rates by two orders of magnitude. Furthermore, the concept of distributed-coupling is extended to superconducting accelerators. Compared to conventional designs, the provided optimization flexibility of the distributed-coupling topology leads to optimized geometries with a reduced surface magnetic field and RF power loss. This reduction should allow for high-gradient operation and reduced system cost. We present our initial attempts to build and test a superconducting distributed-coupling linac. Finally, the concept of distributed-coupling is extended to utilize two accelerating modes that operate simultaneously in the same linac. Dual-mode acceleration enhances the shunt impedance while allowing the structure to operate at much higher gradients. The latter advantage is due to the fact that a given point on the cavity surface does not experience the sum of the peak fields from the two modes at the same time. An extra degree of freedom is obtained by not requiring the operating frequencies to be harmonically related; it is sufficient to have a common sub-harmonic. The value of this sub-harmonic determines the distance between the bunches that can be accelerated. The proposed dual-mode architecture prevents the leakage of the high-frequency mode through the coupling ports of the low-frequency mode by introducing a choke feature in the low-frequency port. Moreover, this architecture preserves the structure symmetry and allows for manufacturing the structure from quadrant copper blocks.