Laser Setup Challenges No-Cloning Rule

Richard Gaughan

The no-cloning rule, a consequence of quantum mechanics, states that no device can produce perfect copies of an unknown quantum mechanical sys-tem. Nevertheless, a system that fairly well copies any possible input state is possible in theory. Now, a team at Oxford University in the UK has demonstrated such a universal cloning device.

The researchers used 120-fs pulses of 780-nm radiation from a Spectra-Physics Inc. Ti:sapphire laser as both the pump and the signal for down-conversion in a BBO crystal. They created the signal photon by splitting off a small part of the beam with a beamsplitter, polarizing it and further attenuating below the single-photon level. The larger part of the beam, after frequency doubling, pumped the 2-mm-thick nonlinear crystal.

An incoming signal photon propagating through the BBO has an approximately 1 in 1000 chance of stimulating emission, and an orthogonally polarized down-converted photon pair also may be spontaneously emitted. In the experiment, one down-converted photon would be emitted in the same direction as the signal photon, and the other would propagate in another direction. The possible measured states, therefore, were two photons emitted in the direction of the signal beam and one in the other direction.

In the absence of stimulated emission, each of two possible states has an equal probability of appearing. But when the time delay between the signal and pump beams is adjusted for coincidence, thus allowing stimulated emission, the state with two identically polarized photons in the direction of the signal becomes twice as likely. When the team considered the probabilities of all possible output states, it calculated a net 83 percent chance that the device would generate a perfect copy for any input polarization state.

The researchers tested the fidelity of the cloning by varying the time delay between the signal and the pump for vertically, 45° and circularly polarized beams. The number of identically polarized photons increased as they had anticipated when they adjusted the time delay for coincidence, and the measured fidelity was 80 percent for the three polarization states.

Antía Lamas-Linares, a graduate student in Dik Bouwmeester's group at Oxford University, said that, although scientists have studied parametric down-conversion extensively as a mechanism for entangling photons, there had been few attempts to use it to demonstrate single-particle cloning. The approach is a natural candidate for a universal cloning machine, she said, because it works by the well-studied principle of stimulated emission and does not require complicated quantum networks.

Bouwmeester's group, which is relocating to the University of California, Santa Barbara, will continue to use the process to study high-order entanglement in photons and hopes to use these states for quantum information processing.