Encryption keys used today rely on the belief that it takes huge computing resources to break them, says Jeffrey Shapiro, of MIT's Optical and Quantum Communications Group. But if someone invents a vastly more powerful quantum computer, that advantage would be lost. In addition, the random sequences of numbers generated to make today's encryption keys aren't truly random. They're generated by mathematical operations, and a smart code breaker might be able to figure out the algorithm being used to generate them. Quantum bits, on the other hand, are completely unpredictable, so the keys based on them should be unbreakable. That's appealing to businesses that want to send financial data securely, as well as to governments, which have all sorts of sensitive communications. "We all know that the security of data is one of the essential issues these days," says Zeilinger.

"I think it's wonderful work," Shapiro says of the European group's paper. "The impressive thing about this is they've done it over such a long distance."

The best that researchers had previously done was to detect entangled photons across distances of about 10 kilometers. To improve on that, Zeilinger's team switched to a laser that emits light in pulses instead of a continuous beam. The pulsed laser only has a repetition rate of 249 megahertz--far slower than the 10 gigahertz lasers commonly used in optical communications networks, which limits how much of a signal can be sent in a given time period. The pulsed laser is also not quite as good as the continuous one at producing entanglement. But it's close, and it gave the team members much more control over when they were producing photons, which helped them separate the photons they wanted from stray light at the detector, so they could read the signal more reliably. The researchers also had to deal with atmospheric turbulence that distorts the photons' path. They used an automated system that continually adjusted the alignment of the telescope to take care of that, although the light beam still wandered over the detector somewhat.

The hope, Zeilinger says, is to improve the lasers and detectors enough that such free-space links work between ground stations and satellites, so that encryption keys can be sent from anyplace on Earth to any other. As most communications satellites orbit at heights of 300 to 500 kilometers, "with our 144 kilometers, we are getting there," he says.

The fact that the team covered that distance in free space "certainly is very significant," says Prem Kumar, director of the Center for Photonic Communication and Computing at Northwestern University. He has sent entangled photons over optical fiber, which is fine for short distances, he says. But because fiber absorbs photons, it's not practical for more than 100 to 200 kilometers, which wouldn't allow for worldwide distribution.

The researchers are part of a European consortium of about 20 groups, called SECOQC, working on secure communications based on quantum cryptography. The consortium aims to test a secure system in Europe sometime next year.