APS News

Chip-scale Accelerators Verge on Breakthrough

By Calla Cofield

At a press conference at the 2013 APS March Meeting in Baltimore, Rodney Yoder of Goucher College predicted that within a year, he and his colleagues will accelerate a beam of electrons through a particle accelerator smaller than the eye of a needle. Yoder showed off samples of his team’s design structure, called a micro accelerator platform, or MAP.

Yoder stated that these new devices could “democratize” accelerator science, and make their use widespread. Small particle accelerators could reduce the need for scientists to use large-scale, multi-million dollar accelerator facilities. Less powerful versions of the device could be used to inspect packages at border checkpoints.

Yoder and co-creator of the MAP structure, Gil Travish of UCLA, first confirmed sending electrons through the structure in results published in 2011. This isn’t a great challenge in an accelerator if the electrons pass through a 2-inch wide copper pipe, but it is a major milestone when the “pipe” is only few microns wide. Their next step, acceleration, is achieved when laser light is injected into the structure, and electrons “surf” on the light waves to gain energy. Yoder says the MAP structure design has the potential to deliver energy gains of 1 GeV per meter, although for many potential applications the structures would most likely stay on the order of about ten centimeters.

The MAP collaboration is based at UCLA, but the structure is undergoing testing at SLAC National Accelerator Laboratory by the Dielectric Laser Acceleration (DLA) Group, which is led by Joel England.

England said in an interview that SLAC scientists have also sent electrons through two other micro-scale accelerator structures, and have discussed those results publicly. The group began tests injecting lasers into all three structures in December of 2011, but spent a year developing test beams that worked for the structure. This required ways of monitoring the very small laser and electron beams and being able to tweak them in various ways. In December of 2012 they began specialized tests of the structures with the ideal beams, and England confirmed Yoder’s anticipation that within the year they will have results showing successful acceleration.

While these results are still preliminary, they represent major gains that as recently as 10 years ago, many people in the accelerator community doubted would ever materialize. Now these accelerator designs are in the running to become the first operational devices to change the paradigm of accelerator size.

“The fact that we’ve gotten particles through them,” said England, “and are actually beginning to test them with lasers…definitely sets the stage for this type of technology as something with a lot of promise.”

These “chip-scale” accelerators are officially called dielectric-based laser accelerators, or DLA’s, in reference to the dielectric materials they are made from, and the use of lasers as an energy source. The MAP structure uses a titanium-sapphire laser which is tunable in the red and near-infrared wavelengths. They operate on the same basic idea as large accelerators: Bunches of particles gain energy by “surfing” on electromagnetic waves.

The particle bunches gain energy over distance, which partly accounts for the size of the world’s most powerful accelerators, like the 2-mile-long linear accelerator at SLAC or the 27-km Large Hadron Collider. Accelerator size is also determined by the wavelength of light used to boost the particles. The particles pass through a central “pipe,” and light waves are injected all along the length of the pipe, through cavities that run perpendicular to it. These cavities need to be roughly the wavelength of the light used. Most large accelerators use microwaves, meaning each cavity needs to be a few centimeters wide. To use optical light requires cavities on the scale of a few hundred nanometers to a few microns.

While designs for optical-light accelerators date back as far as the 1960’s, two breakthroughs in the last 15 years have made them much more feasible. The cost of optical lasers continues to drop, including the cost of fiber lasers, which Travers says will work as an energy source for the DLA’s. At the same time microscale fabrication techniques, driven mostly by the semiconductor industry, have also improved.

DLA designs differ mostly in their architectures. Yoder and Travish’s design is called a “slab” structure, because it consists of two flat slabs of dielectric material. Sandwiched between the slabs is a section of vacuum where the particles pass through. The pair began serious work on the design in 2006, having both worked with James Rosenzweig at UCLA, though at different times.

The particle bunch duration  from a DLA structure is only a few attoseconds, which opens up the possibility of imaging very fast processes, because the bunch duration functions like a shutter speed. The bunches may contain as few as 100 to 1000 electrons each, resulting in a low charge per bunch and a weak signal to a detector, so they may be  grouped together into picosecond-long “macrobunches,” that provide significantly more charge and signal. The beam may then produce over a million of those macrobunches (over a billion individual bunches) per second. Yoder and England both estimated that in five years, they would have a test accelerator. This would mean the ability to line up multiple structures in series, and accelerate a beam through them, to observe a higher energy gain.

There is still a significant amount of work to be done before a prototype accelerator could be used for experimentation. Neither of the research groups have an electron beam small enough to fit exclusively through the beam “pipe” of any of these microscale structures. For their tests they have used a beam that is larger than the structure’s beam tunnel, and had to confirm that portions of the beam did pass through it. In addition, development of a full prototype will require detectors for these devices and diagnostic tools to study the beams they produce.

“I don’t know if one research group is capable of doing that on its own,” said England. “So we have collaborators that are experts on [semiconductor] fabrication techniques…and experts on lasers, experts on materials science. We’re trying to fold new people into the field to help tackle some of these challenges.”

APS encourages the redistribution of the materials included in this newspaper provided that attribution to the source is noted and the materials are not truncated or changed.

Editor: Alan Chodos
Staff Science Writer: Michael Lucibella