Miniature cyclotron built at LBL

June 17, 1994

Lynn Yarris,

BERKELEY, CA -- In answer to the current fiscal pressure on science to scale back, a cyclotron the size of a microwave oven has been developed by researchers at the Lawrence Berkeley Laboratory.

Though comparable in size to the early cyclotrons of LBL-founder and namesake, Ernest Lawrence, this new miniature cyclotron is not meant to do nuclear physics, but is instead a tool for measuring trace amounts of elements and their isotopes.

For the past three years, Anthony Young, a physicist in LBL's Accelerator and Fusion Research Division (AFRD) has been leading the effort to make a practical, laboratory-sized cyclotron for accelerator-based mass spectrometry. This highly sensitive technique, which can cull and identify single atoms from a sample containing trillions, can be used for a variety of purposes, such as radiodating fossils and archeological artifacts, measuring atmospheric pollutants, sequencing DNA, or studying the metabolism of therapeutic drugs.

Young and his colleagues credit Rich Muller, an astrophysicist in LBL's Physics Division for having launched the modern era of accelerator-based mass spectrometry in 1976 when he developed it as an improvement over radioactive carbon dating.

"We think of Rich Muller as the godfather of our project," says Young. "He was the one who thought to combine mass spectrometry with accelerator technology as a way to make radiodating much more sensitive."

In conventional radiodating, the age of a sample is estimated by taking a count of the one or more carbon-14 atoms that are decaying to nitrogen. Carbon-14 is a radioactive isotope with a half-life of 5,370 years whose proportion to stable carbon-12 is constant in all organic matter.

In accelerator mass spectrometry, the total number of carbon-14 atoms in the sample are counted whether they are decaying or not. This results in a many thousandfold increase in sensitivity,

which means either a much smaller amount of sample is needed, i.e., a tenth of a milligram instead of a gram, or else detection time can be greatly shortened.

Accelerator mass spectrometry as practiced today is remarkably precise but has not been widely adapted because of the size and expense of tandem Van de Graff accelerators, the machines of choice for radiodating samples. (Less than two dozen facilities perform AMS in the world today.)

Muller and his research group did build a couple of prototypes for a low-energy, relatively inexpensive miniature cyclotron -- dubbed the "cyclotrino" -- that, like tandem accelerators, were well-suited to measure the presence of carbon-14. Several other versions were also built by groups outside of LBL. In all cases, however, a variety of technical difficulties hampered further development.

"The project became such an engineering tour de force that many people came to believe it couldn't be done," says Young. He and his colleagues have made some crucial improvements in their mini-cyclotron that should rekindle interest in the technology. Perhaps the most important improvement has been the replacement of an electrical magnet with a series of permanent magnets to generate the cyclotron's magnetic field.

In accelerator mass spectrometry, a sample is vaporized into ions which are then accelerated and sorted according to mass and charge as they pass through a magnetic field. A particle detector is then used to identify the ions. By generating the magnetic field with permanent magnets instead of the electrical magnets used in previous compact cyclotrons, Young and his colleagues reduced the overall size and weight of their machine and minimized its power and cooling demands. Permanent magnets also give this new mini-cyclotron a portability that older versions lacked.

There was a question as to whether permanent magnets could provide a sufficiently uniform field. However, by arranging approximately 200 individual samarium cobalt magnets (designed by Engineering Division's Klaus Halbach and Ross Schlueter) around a steel yoke that features a fixed gap of 16 millimeters between its poles, Young and his colleagues were able to provide a highly uniform 1.0 Tesla strength magnetic field through which ions can pass in an isochronous orbit.

"Magnetic fields are not so critical in conventional accelerator mass spectrometry where they are only used in the analyzer," says Young. "In cyclotron mass spectrometry, however, where magnetic fields are integral to the acceleration process, if the field is not uniform, the technique will fail."

Another major improvement was in the injection line that delivers a beam of accelerated ions from its source to the cyclotron's magnetic field. Other compact cyclotrons used a radial injection system that is straightforward but reduces spectrometric sensitivity.

"We designed a spiral inflector with an electrostatic channel that twists or tilts as it guides the ions down the axis of the machine and into the midplane of the magnetic field," says Young. "This axial injection of the ions improves sensitivity and increases sample throughput."

Sensitivity and detection efficiency were further improved by the use of a negative ion source of the type designed by AFRD's Ka-Ngo Leung for fusion research. With its unique multicusp magnetic field (the name comes from the geometry of the magnetic lines of force), the negative ion source is not only much easier to operate than the sputter sources used in previous mini-cyclotrons, it also yields a superior quality beam at a much higher current and sample efficiency. A higher current means there are more ions in the beam available for measurement.

Like the tandem accelerators and other compact cyclotrons, the new minicyclotron has been optimized to sort and identify carbon-14 ions. A problem with this has always been that there are several other isotopes and molecules, for example, nitrogen-14 or carbon deuterium, that have almost an identical mass but are much more plentiful than carbon-14. With its heightened sensitivity, the new minicyclotron can make these distinctions.

"Our minicyclotron is not a competitive technology to conventional accelerator mass spectrometry, but an alternative for those who can't afford to use a tandem accelerator," says Young. "Building a machine like this is something that could not be done at many other places because its design and fabrication called for so many different skills."

Young's collaborators, in addition to those already mentioned, included Kirk Bertsche, a former member of Muller's group, Dave Clark of the Nuclear Science Division, and AFRD's Wulf Kunkel, Chaoyang Li, Mary Stuart, and Russ Wells.

LBL is a U.S. Department of Energy national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California.