Axion gone: new tests find no sign of anomalous particle.

Last year, physicists reported seeing tantalizing experimental traces of the axion, a hypothetical subatomic particle that's been mentioned as a possible constituent of cosmic dark matter. But the axion was showing up where theory said it shouldn't be. It now looks as if it wasn't there after all.

The axion sprang from an attempt to explain certain differences between the strong and weak nuclear forces. Cosmologists seized on the axion because its properties made it a plausible component of dark matter, the unseen material that far outweighs ordinary matter in the universe.

In 2000, Giovanni Cantatore and his colleagues at the Italian National Institute of Nuclear Physics in Legnaro were investigating the behavior of photons by shining a laser beam through a strong magnetic field. They noticed that the light's polarization shifted slightly after it went through the field--not the effect they were looking for.

The team posited that the polarization shift could have resulted from the magnetic field converting some of the beam's photons into axions, which would then fly off undetected. But the shift the researchers saw, while tiny, was much larger than physicists had thought possible. If such an effect occurred in the cores of stars, for example, axion emission would siphon energy away, reducing stellar lifetimes far below their actual values.

Cantatore and his colleagues reluctantly decided to publish their data last year, after numerous fruitless efforts to find a flaw in their experiments. "We thought it was our duty to report our results,' Cantatore says.

After the announcement, at least five labs around the world began experiments to settle the issue. They looked for a different effect, known as photon regeneration, or, in Zenlike fashion, as "light shining through a wall." Researchers shoot a laser beam through a magnetic field toward a metal plate. The metal wall blocks photons, but any axions created in the field would pass through. On the other side of the wall lies a second magnetic field that would convert some of the axions back into photons, making it appear that some photons had passed through.

A team at the Ecole Polytechnique in Palaiseau, France, reports the first results of such an experiment in an upcoming Physical Review Letters. "No regenerated photons were observed," says team member Cecile Robilliard, of the Universite Paul Sabatier in Toulouse, France. "This allows us to exclude the Italian results with 99.9 percent confidence," she says.

Meanwhile, Cantatore and his colleagues have performed a new round of observations after taking their machine apart and rebuilding it almost from scratch. The polarization shift finally went away. The team posted a retraction of its earlier results online last June.

Cantatore says that a number of small effects could have combined to create the fake signal. For example, magnetic field lines might have leaked out of the magnet and helped shift the polarization.

Helen Quinn, a theorist at the Stanford Linear Accelerator Center in Menlo Park, Calif., who helped propose the axion in the 1970s, says that other experimental approaches might still find axions in the future.