Flipping a quantum mechanical coin.

Physicists generally regard the emission of light by an excited atom as a random process. They assume it's impossible to predict precisely when one of the atom's electrons will drop from a higher to a lower energy level, emitting a photon in the process. But this randomness doesn't follow automatically from the other fundamental principles of quantum theory, says Thomas Erber of the Illinois Institute of Technology in Chicago.

To test for randomness, Erber analyzed data obtained from Wayne M. Itano, David J. Wineland, and their coworkers at the National Institute of Standards and Technology (NIST) in Boulder, Colo. The NIST researchers had studied light absorbed and emitted by a single mercury ion held in a special trap and irradiated continuously by laser light of a certain frequency. In response, the suspended mercury ion flashed on and off at irregular intervals, acting like an optical telegraph.

Applying a variety of techniques developed in recent years to search for patterns in encrypted data and other strings of digits, Erber has studied a sequence of 20,000 numbers representing the time intervals between successive quantum jumps in the NIST data. So far, no discernible patterns have emerged. To be sure of his result, however, Erber needs to check longer sequences of numbers. If no patterns are found, the quantum jumps of a single ion may prove an infinite source of "cryptographically invulnerable" random numbers, Erber notes.