A schematic of the polarised electron source for the ILC.

Imagine that you are given roughly three months to complete a project. Then your boss comes in and asks if you can speed things up – not by just a couple of weeks or even months. You have precisely 100 times less than your original deadline, squeezing the amount of time in three months down to one day. Picture yourself in this situation, and you will have a renewed respect for the massive requirement of the electron source R&D group for the International Linear Collider – shortening the electron pulses from two nanoseconds to 20 picoseconds.

Generating and shortening bunches for the ILC polarised electron source is exactly how Christian Haakonsen, an undergraduate at McGill University in Montreal, Canada, spent his summer at SLAC as part of the Department of Energy's Science Undergraduate Laboratory Internship (SULI) program. Undergoing possibly the shortest crash course in accelerator physics, Haakonsen worked closely with SLAC physicist Axel Brachmann to learn the intricacies of the baseline design for the electron source. "This is a whole new field to me," Haakonsen said. "I had never done anything like this before, and I am very pleased with the experience."

Consisting of a 120 kilovolt direct current gun, the electron source in the ILC will generate electrons by shining a laser into a photocathode. Producing polarised electrons for the high energy beam (meaning that they will all spin in one direction) is just one piece of the puzzle though. In order for the electrons to fit inside the superconducting cavities in the ILC, they must be compressed to extremely small sizes, requiring the pulses in the electron source to be 20 picoseconds.

Twenty-two SULI students had internships at SLAC this summer. Christian Haakonsen (far right) spent his 10-week internship testing the baseline design for the electron source in the ILC.

Based upon a twenty-year-old SLAC design that generates two-nanosecond pulses, the electron source in the ILC will use cavities called bunchers to shorten the pulse-length down to the required 20 picoseconds. While this specific ILC electron source work has been done on a conceptual level, details in the design need to be refined in order to produce the highest quality electrons. Using a combination of three software programs, General Particle Tracer, EGUN and Superfish, Haakonsen simulated the generation and initial behavior of the electrons.

"We increased the realism in the simulations by adding certain elements such as a magnetic field," he said. By adding some bending magnets to the equation, Haakonsen was able to test how the electrons would react to such conditions in the real machine. "We were able to do it without compromising performance," he said.

Haakonsen also determined areas that possibly need to be reworked in the design, such as the configuration of the RF bunchers. "Right now we use two 5-cell RF bunchers, but the incoming electrons don't have enough energy," Haakonsen said. "They end up not compressing and expand instead. We need to go back to see if we can replace the 5-cell buncher with something else."

Haakonsen still has another year left at McGill before pursuing graduate school, but he already expressed interest in accelerator physics, especially with the satisfaction of making such real contributions to the ILC. "When I started the internship, I was hoping to improve the design," he said. "But I found something unexpected that needs to be changed. It is good to find something like that at this early stage, and it is what makes simulations so useful."

-- Elizabeth Clements